Tools to Facilitate Circuit-Mapping Using the Split Gal4 System. William C Shropshire1, Haojiang Luan1,2, Benjamin White1. 1) Section on Neural Function, NIH, NIMH, Bethesda, MD; 2) Janelia Farm Research Campus, Ashburn, VA.

   To identify neurons that function in behavioral circuits, Gal4 enhancer trap lines are used to drive UAS-effector transgenes that alter neuronal activity. If manipulation of neural activity results in a behavioral change, all or part of the circuit must lie within the Gal4 expression pattern. Refining such patterns to determine which cells specifically belong to the circuit typically requires use of intersectional methods: e.g. Gal80, or the Split Gal4 technique, in which the Gal4 DNA Binding Domain (i.e. Gal4DBD) and a transcription activation domain (e.g. VP16AD) are independently targeted to different subsets of cells. The incompatibility of Gal80 and the Split Gal4 system has limited their combined use to achieve further refinement. Here we introduce an inhibitor of Gal4DBD function, called the Killer Zipper, which functions like Gal80 in the Split Gal4 system. In addition we introduce an enhancer trap Gal4 construct (i.e. GG) that can be used with Gal80 and then converted in vivo to a Gal4DBD construct using cre recombinase. The Killer Zipper consists of the Gal4DBD fused to the same heterodimerizing leucine zipper that is fused to the VP16AD. It therefore inhibits transcriptional activity by competing with VP16AD for binding to the Gal4DBD and also by promoting formation of Gal4DBD homodimers, which can, in theory, bind UAS sites and prevent the binding of productive Gal4DBD-VP16AD dimers. We have demonstrated the efficacy of the Killer Zipper expressed in a group of neurons that express the neuropeptide CCAP. We have also generated enhancer trap lines with the GG construct and shown that the floxed Gal4 component can be excised in vivo using a heat-shock cre recombinase. Three GG lines have been converted from Gal4 to Gal4DBD expressing lines. We anticipate that the Killer Zipper and GG systems will be useful in improving the resolution of circuit mapping neuronal screens.