Project Summary T Stellate cells of the ventral cochlear nucleus (VCN) form an important ascending pathway that transmits spectral information from the auditory nerve to numerous auditory nuclei. They innervate the olivocochlear efferents in the ventral nucleus of the trapezoid body, the lateral superior olive, the inferior colliculi and the thalamus. In preliminary experiments we have discovered that groups of T stellate cells within an isofrequency lamina are bidirectionally interconnected through excitatory synaptic connections that can be potentiated. In dual, whole-cell patch-clamp recordings from T stellate cells, firing in a presynaptic cell generally evoked no EPSCs in the postsynaptic cell unless presynaptic firing was paired with postsynaptic depolarization. These findings are exciting for two reasons. First is that the mechanism underlying that potentiation is new and unprecedented. Postsynaptic depolarization increased the probability of recorded EPSCs, a presynaptic function, implicating the involvement of a retrograde messenger. Our preliminary results support the hypothesis that nitric oxide serves as that retrograde signal. Aim 1 is to use intracellular recordings in slices to gain a deeper understanding of the mechanisms that underlie potentiation of connections between T stellate cells and to understand their source and dynamics. We will identify what neurons participate in polysynaptic connections, how synaptic excitation by auditory nerve fibers affects the plasticity of interconnections, examine signaling through the nitric oxide pathway, and measure rates at which potentiation develop and fade. Second is that our discovery reveals a new form of central gain control at the network level. Bidirectional, excitatory interconnections indicate that T stellate cells in an isofrequency lamina form a network and could explain how T stellate cells can sharpen the encoding of spectral peaks. These interconnections could also form synaptic positive feedback loops that lead to hyperexcitability in the face of loss of auditory nerve fibers and the consequent uncoupling of excitation and inhibition. Aim 2 is to use computational neural models to understand the implications of excitatory interconnections between T stellate cells on their encoding of sound. We will implement models that can simulate the response features of single T stellate cells and build an interconnected neural network to understand how network connectivity contributes to potentiation. We will test the hypothesis that excitatory interconnections enhance the encoding of spectral peaks and that inhibition is required to stabilize the network.