Neurobiology

Dietmar Schmucker, Ph.D.

Associate Professor of Neurobiology
Harvard Medical School
Department of Cancer Biology
Dana-Farber Cancer Institute

My laboratory uses the model organism Drosophila (fruit fly) to study molecular mechanisms that control the development of neuronal connectivity. Any assembly of neuronal circuits, simple or complex, is controlled by a series of specific molecular signaling systems that instruct the directional growth of neuronal processes (axons and dendrites). We combine genetic, biochemical and cell biological approaches to study how neuronal surface receptors control these signaling systems during axon and dendrite guidance, branching and synaptic target selection. We focus our developmental and functional studies primarily on the analysis of sensory neurons of the visual and somato-sensory system of flies.

One important goal is to understand how a relatively small number of neuronal signaling systems can achieve the exquisite specificity that is required to specify millions of synaptic connections.

We have identified a Drosophila protein (Dscam) that is functionally required for the proper formation of nerve connections in the CNS and PNS (Schmucker et al. 2000). Dscam is specifically expressed on the surface of growing nerves. Dscam is a member of the immunoglobulin superfamily and highly related to the human protein Down Syndrome Cell Adhesion Molecule (DSCAM). The Drosophila Dscam gene is highly complex. Through alternative splicing some 38,000-protein isoforms can be formed. All Dscam isoforms share the same overall molecular architecture but differ in four protein domains. Such an extraordinary molecular diversity has previously not been described for any other receptor expressed in the nervous system but is reminiscent of immunoglobulin receptors in the immune system.

To address the functional requirement of Dscam we are using genetic methods and single cell labelling to analyze pathfinding, targeting and synaptogenesis of mechanoreceptor neurons (ms-neurons) in the somatosensory system of adult flies. We are analyzing topographically organized ms-neuron projections into the CNS. Our phenotypic analysis of Dscam null alleles showed that Dscam is essential for proper axonal branching and targeting of ms-neurons. In addition, we have generated flies that lack a substantial amount of the Dscam receptor diversity but maintain normal amounts of Dscam protein. Remarkably, animals that have an isoform repertoire of maximally 22,000 instead of 38,000 receptors already display distinct defects in neuronal targeting. These experiments provide evidence that a very large number of structurally unique receptor isoforms is required to ensure fidelity and precision of neuronal connectivity (Chen et al. 2006). At present we are conducting biochemical and genetic experiments addressing the mechanistic aspects of how receptor diversity might be utilised during synaptic targeting in the somatosensory system.

Many studies (primarily in mammals) have revealed that receptor functions and signal transduction mechanisms are remarkably similar between the nervous system and the immune system. Furthermore diverse immune receptors are expressed and required for neuronal wiring (reviewed in Boulanger, Huh and Shatz, 2001). Remarkably, we found recently that the hypervariable neuronal receptor Dscam is also expressed in the immune system of flies. In fact, Drosophila immune-competent cells have the potential to express more than 18,000 isoforms of the Dscam receptor (Watson et al. 2005). Dscam isoforms can differentially bind to bacteria and hemocyte-specific loss of Dscam impairs the efficiency of phagocytic uptake of bacteria. Generally, our findings suggest an unsuspected molecular complexity of the innate immune system of insects. We are therefore interested to elucidate general mechanisms that underlie receptor specificity by comparing neuronal and immunological signaling systems. To this end we are using a series of different approaches (biochemistry, X-ray crystallography, microarray analysis, and genetic analysis) to determine similarities as well as differences of Dscam signaling in different cell types.

Dscam is a member of the immunoglobulin superfamily and contains 10 Ig-domains, 6 fibronectin type III (FNIII) domains, a single transmembrane domain, and a novel 374 AA containing cytoplasmic domain. The Dscam mRNA comprises 24 exons. Tandem arrays of alternative exons 4, 6, 9 and 17 allow the generation of variable Ig domain sequences for Ig 2, 3 &7 and two different transmembrane domains. Mutually exclusive splicing occurs for exons 4, 6, 9, and 17. Genomic and cDNA analysis revealed the potential of generating some 38,000 isoforms of Dscam. Variable exons are shown in color. Gray lines in genomic DNA and boxes in mRNA represent constant exons.