The goal of this research is to exploit techniques for the targeted manipulation of neural activity to identify, and functionally define, brain networks underlying specific behaviors. As a model for such investigations, we are identifying the networks that govern the behavioral program executed by adult fruit flies shortly after emergence from the pupal case. This program consists of an adaptive behavioral phase, which mediates the search for a suitable environment, and an innate phase, which drives expansion of the wings to make them flight-worthy. Elucidation of the circuits underlying both components of this program promises a detailed understanding of how intrinsic and extrinsic factors act, individually and in concert, to recruit motor patterns to assemble behavioral sequences. Identification of the mechanisms by which neuronal networks interact and adapt to organize behavior should shed light on the failures in behavioral organization, which lie at the root of many mental disorders.? ? Our work thus far has focused on the innate component of the post-emergence behavioral sequence in Drosophila. This component consists of two coordinately executed motor patterns, both of which are governed by the hormone bursicon. Using techniques that permit the targeted suppression of specific brain cells in living, behaving flies, we have identified two functionally distinct groups of bursicon-expressing neurons. One group secretes bursicon into the hemolymph (i.e. blood) to effect physiological changes at the level of the wing, while the other secretes bursicon within the central nervous system to induce the motor patterns responsible for wing expansion. Both groups are necessary for wing expansion. In addition to these two groups, we have also identified a group of regulatory neurons that do not express bursicon, but modulate the release of the hormone from the bursicon-expressing groups. ? ? Our work depends heavily on genetic techniques that allow us to manipulate neuronal activity in specific subsets of cells. In the past, we have developed tools for the constitutive silencing (i.e. UAS-EKO) and enhancement (i.e. UAS-NaChBac) of electrical activity in neurons. We have recently supplemented these tools with one that permits the acute activation of neurons using small decrements in temperature (i.e. UAS-TRPM8). This tool, which exploits a cold-sensitive ion channel from mammals, has recently allowed us to determine that the bursicon-expressing neurons of Drosophila are not only necessary for wing expansion, but are also sufficient for this process: Activation of the bursicon-expressing cells initiates the entire wing expansion program. This implies that some or all of the bursicon-expressing neurons act as command neurons for this program.? ? To identify which bursicon-expressing neurons might act as command neurons, we have required more refined techniques for manipulating the function of individual cells, or groups of cells in vivo. To that end we developed the combinatorial, Split Gal4 system, which permits selective gene targeting within a cell group of interest. Split Gal4 incorporates technology from the yeast two-hybrid system in that it divides the Gal4 molecule into its component DNA-binding (DBD) and transcription activation (TA) domains. Each domain is fused to one of two complementary, heterodimerizing leucine zippers so that the DBD and TA domains associate in cells that express both. In these cells, and in these cells alone, is Gal4 transcriptional activity reconstituted. By independently targeting the two domains in vivo, we can activate transgenes downstream of Gal4s UAS binding site selectively in the subset of cells that expresses both domains. We have exploited this system by targeting the DBD domain to bursicon-expressing neurons and making TA enhancer trap lines that express the Gal4 TA (or the more potent TA of the HSV-1 VP16 transcription factor) in arbitrary patterns that include different subsets of the bursicon-expressing neurons. We have generated lines that permit expression of UAS-transgenes in numerous unique subsets of bursicon-expressing neurons and are currently analyzing the consequences of selectively suppressing or activating these neurons. ? ? Investigation of the neuronal substrates of posteclosion behavior in Drosophila using the broad palette of tools we are developing should provide insight into the principles used by all nervous systems to generate and organize behavior. In addition, it should serve as a proof of concept of a circuit mapping approach that can be extended to studies of mammalian behavior as similar tools become available for vertebrate organisms.
