Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Brain and Cognitive Sciences, 2010.Cataloged from PDF version of thesis.Includes bibliographical references.This thesis is a collection of manuscripts addressing connectivity of neural circuits in cultured hippocampal neurons. These studies begin with an investigation of dopaminergic modulation of excitatory synapses in small circuits of neurons grown on glial micro islands. We found that dopamine transiently depressed excitatory synaptic transmission. Scaling up to larger circuits of neurons proved more challenging, since finding connected pairs became combinatorially more improbable. The discovery and use of light-activatable ion channel channel rhodopsin-2 (ChR2) promised to revolutionize the way in which we could map connectivity in vitro. We successfully delivered the gene for ChR2 in hippocampal cultures using recombinant adeno-associated virus and characterized the spatial resolution, as well as the reliability of stimulating action potentials. However, there were limitations to this technique that would render circuit maps ambiguous and incomplete. More recently, the engineering of rabies virus (RV) as a neural circuit tracer has produced an exciting method whereby viral infection can be targeted to a population of neurons and spread of the virus restricted to monosynaptically connected neurons. We further investigated potential mechanisms for previous observations which claim that RV spread is restricted to synaptically connected neurons by manipulating neural activity and synaptic vesicle release. We found that RV spread increased for blockade of synaptic vesicle exocytosis and for blockade of neural activity. The underlying premise for pursuing these methods to elucidate connectivity is that the computational power of the brain comes from changeable, malleable connectivity and that to test network models of computation in a biological brain, we must map the connectivity between individual neurons. This thesis builds a framework for experiments designed to bridge the gap between computational learning theories and networks of live neurons.by Jennifer Lynn Wang.Ph.D.