Integrating information from different senses is critical to navigating the world around us. Many everyday tasks, like reaching for a ringing phone, require attributing features from different senses to the same object, and psychophysical studies show that a variety of species, including both primates and rodents, optimally combine information from different senses in the face of noise. How the brain achieves these feats remains to be understood. Primary sensory cortical areas, which represent the first stage of cortical information processing, have long been considered uni-sensory. An emerging alternative view, however, holds that even primary sensory cortical areas participate in multi-sensory integration. Recent work has shown that a variety of multi-sensory interactions occur even in primary sensory cortical areas, and theoretical studies suggest that networks of reciprocally-connected sensory areas may be useful for performing certain multi-sensory computations. The empirical literature, however, has thus far ignored the role of learning in multi-sensory integration in primary sensory cortical areas, which theoretical studies suggest may be critical. The goal of this proposal is to determine the effect of learning on multi-sensory interactions in primary sensory cortical areas. We will first test the hypothesis that mere experience of cross-sensory statistics modifies multi-sensory responses in primary sensory cortical areas. To test this hypothesis, 2-photon calcium imaging will be used to record neural responses to a panel of somatosensory and auditory stimuli in primary somatosensory cortex (S1) of awake mice. One auditory stimulus will then be repeatedly paired with one whisker stimulus over the course of several days, after which responses to the full panel of stimuli will be recorded again. In particular, repeated pairing may cause S1 responses to the paired tone to more closely resemble S1 responses to the paired whisker stimulus. We will also test the hypothesis that repeated pairing of stimuli of different sensory modalities has an enhanced effect on multi-sensory activty in primary sensory cortical areas when the paired stimuli predict reward. These experiments will expand our understanding of how the brain performs multi-sensory integration and contribute to a significant revision of the traditional view of primary sensory areas as as discrete, parallel modules responsible for a single sense modality each, instead expanding our understanding of how primary sensory areas across the brain can coordinate closely to perform certain computations as a whole.
This project aims to examine the role of primary sensory cortical areas in multi-sensory integration in the healthy brain. Congenitally blind patients who have vision surgically restored in late childhood are unable to visually recognize objects previously known only by touch, suggesting learned associations are necessary for hetero-modal feature binding. This work will hopefully elucidate how these associations are learned at a neural circuits level and thereby inform future treatments for deficits in multi-sensory integration resulting from conditions like congenital blindness and amblyopia.
