Despite impressive advances in almost every field of neuroscience, our understanding of brain function remains largely confined to its building blocks at the microscopic level, and to phenomenological descriptions at the macroscopic level. Transgenic mice have provided neuroscience with countless correlations between microscopic and macroscopic levels, but because most gene knockouts have lacked regional and temporal specificity, insights into computational mechanisms remain scarce. Understanding how comp lex mental functions originate from electrical and chemical processes in brain cells requires an integrated systems-biological analysis focused on the intermediate level of neuronal microcircuits, where myriads of intricately connected neurons with differ ent properties act together. The greatest challenge for an intermediate-level analysis is the need to restrict genetic interventions in freely moving animals to particular brain regions and cell-types. Technologies for this purpose have been developed onl y recently. The aim of the present project is to exploit and improve these recent developments to understand computation in a prototype non-sensory cortex. Using the rat spatial representation system as an experimental model, we shall introduce molecular markers into specific cell types of the hippocampus and entorhinal cortex by stereotaxic injection of viral vectors. Viral vectors shall also be used to introduce transgenes that silence, activate or modify the target cells by interfering with their membr ane potential or their synaptic release machinery. Temporal specificity will be induced by expressing exogenous light-activated channels in target cells in rats implanted with a fiber optic probe connected to a laser source. Stimulation at specific wavele ngths can activate the channel protein at a time scale of milliseconds. This approach provides entirely new opportunities for determining the function of specific components of the hippocampo-entorhinal system.