The molecular basis of neural computation.
A central goal of our research is to identify the cellular and molecular mechanisms that shape the physiological properties of neurons and how these properties enable them to effectively process visual information. Using Drosophila as a model organism offers the ability to specifically manipulate molecular function in any cell type of interest using genetic tools. We combine genetic manipulations with in vivo 2-photon calcium imaging or with the analysis of visually evoked behavior. Using this approach we aim to determine the molecular mechanisms that are important for the processing of visual signals, and link these mechanisms to wider circuit function and behavior.
Cell type specific genetic tools for circuit analysis
The analysis of neural circuits or the molecular machinery of neural computation relies on genetic access to individual cells or cell types, or distinct access to two potential synaptic partners. Several efforts have recently been made to develop genetic tools that allow this degree of specificity. Starting in the Clandinin lab at Stanford University, USA, we (Gohl, Silies et al. 2011) have developed a genetic toolkit, the InSITE system, that can be used to refine and repurpose expression patterns. To this end, we have generated ~2000 InSITE Gal4 driver lines that can be genetically replaced in vivo with components of e.g. the splitGal4 system for genetic intersections, or components of other binary expression for distinct genetic access to different cell types. The toolkit as well as the full Gal4 driver collections are publicly available. We are continuing to develop tools for more specific manipulations of subcellular circuit elements.
Neural pathways that process visual information
To extract visual motion, the nervous system must compare signals over space and time. How this could be achieved, was solved computationally decades ago. Since then, visual motion-detection has served as a premier context in which to investigate how the nervous system performs specific computation. While the core elements of motion detecting circuits have recently been proposed, it is already clear that specific aspects of motion computation are more distributed than previously anticipated and that even the peripheral processing of visual information in the fruit fly uses relatively complex circuits. How distributed are neural computations in the fly visual system? What is the role of the parallel pathways that we find? Do they help the animal to handle different behavioral or environmental contexts or adapt to different scene statistics, or are they merely making the system more robust?
Our goal is to disentangle these complex visual circuits and define the function of individual neurons, both in terms of their physiological specialization as well as their role in guiding visual behaviors.