We study two key questions in neurobiology: (1) how neurons form complex connections or circuits during development and (2) how the assembled neuronal circuits function to guide animal behaviors. We use Drosophila color vision as a model to study these questions. The fly retina contains three types of photoreceptors, R1-6, R7, and R8, each responding to a specific spectrum of light and connecting to a specific layer in the brain. We have been focusing on the layer-specific connections made by the UV-sensitive R7 neurons. Using wavelength selection behaviors, we isolated several mutants in which various aspects of R7–brain connectivity are affected. We are currently cloning and analyzing these mutants. To study the function of color-vision circuitry, we focus on the identification of the medulla target neurons that serve as synaptic targets for photoreceptor neurons.
Molecular mechanisms regulating target selection of R7 photoreceptor axons
Yonekura, Ting, Gao, Xu; in collaboration with Chess, Chiba, Hsu, Neves
Using single-cell mosaic techniques, we first determined the developmental processes of R7 target selection. We found that R7 axons select their target layer in two distinct stages. During the late larval and early pupal stage (the first target-selection stage), R7 neurons sequentially differentiate and project axons into the R7–temporary layer, where they remain for one to two days. During the late pupal stage (the second target-selection stage), all R7 growth cones regain motility and synchronously project into the destined layer, the R7–recipient layer. The characterization of the development of R7–brain connections provides a framework for studying the isolated mutants.
In a genetic screen based on wavelength selection behavior, we identified several mutants that have defective R7–brain connections. The mutants include ones in two known genes, N-cadherin and LAR, as well as two yet-to-be-determined loci, pex (premature extension) and ovs (overshoot). The removal of N-cadherin from R7 axons results in the mutants’ failure to extend into the R7–temporary layer at the early pupal stage; subsequently at the adult stage, they misconnect to the R8–recipient layer. Drosophila N-cadherin belongs to a unique family of evolutionarily conserved classical cadherins that have large complex extracellular domains and catenin-binding cytoplasmic domains. To determine the functional domains of N-cadherin, we conducted a structure-function analysis. We found that the cytoplasmic domain of N-cadherin is not essential for mediating homophilic interaction in cultured cells and is largely dispensable for layer-specific targeting of R7 axons in vivo. However, the N-cadherin cytoplasmic domain, and hence its catenin-binding activity, is required to maintain proper morphology of R7 growth cones. Domain swapping with the extracellular domain of N-cadherin2, a related but nonadhesive cadherin, revealed that the N-cadherin extracelluar domain is required for both adhesive activity and R7 targeting. Together, these results suggest that N-cadherin mediates adhesive interactions and not cytoplasmic signaling to regulate R7 target selection.
By analyzing N-cadherin isoforms generated by naturally occurring alternative splicing, we examined how variable sequences in the extracellular and transmembrane domains affect N-cadherin’s adhesive activity. We found that all N-cadherin isoforms mediate homophilic interactions but that the isoforms encoded by exon 18b have a higher binding activity than those encoded by the alternative exon 18a. Domain-swap experiments further revealed that the different sequences in the transmembrane domain among isoforms are responsible for the isoforms’ differential homophilic binding activity. Using quantitative PCR analyses, we found that the alternative splicing of exons 18a and 18b is developmentally regulated and thus might provide a means to fine-tune N-cadherin adhesive activity at different developmental stages.
The mutations pex and ovs affect R7 target selection in the opposite way to that of N-cadherin and LAR mutations. In contrast to N-cadherin and LAR mutant R7 axons, which retract to the superficial R8–recipient layer, pex and ovs mutant R7 axons overshoot the target region. Interestingly, ovs mutant R7 axons overshoot the R7–recipient layer and terminate at a deeper layer while pex mutant R7 axons extend into the neighboring target area occupied by other R7 termini or loop back to more superficial layers. Developmental analysis revealed complex pex phenotypes: pex mutant R7 axons overshoot the R7–temporary layer at the early pupal stage but retract to the R7–temporary layer at the mid-pupal stage. However, during the late pupal stage, pex mutant R7 growth cones regain motility and extend into the neighboring target region. We hypothesize that pex and ovs provide the opposing (or balancing) effect on the R7 growth cones, presumably by counteracting N-cadherin or LAR function. We are currently cloning pex and ovs in the hope that their molecular identity will provide insight into the regulatory mechanisms of R7 target selection.
Lee RC, Clandinin TR, Lee C-H, Chen PL, Meinertzhagen IA, Zipursky SL. The protocadherin Flamingo is required for axon target selection in the Drosophila visual system. Nat Neurosci 2003;6:557-563.
Ting CY, Yonekura S, Chung P, Hsu SN, Robertson HM, Chiba A, Lee CH. Drosophila N-cadherin functions in the first stage of the two-stage layer-selection process of R7 photoreceptor afferents. Development 2005;132:953-963.
A directed mosaic technique for manipulating medulla target neurons
Chung, Yonekura
Owing to the available genetic mosaic methods that allow genetic manipulation of R7 neurons and visualization of mutant R7 axon projections at the single-cell resolution, we identified several genes required presynaptically in the R7 neurons for target selection. However, similar tools were not available for R7–target neurons in the medulla ganglion. As a result, little is known about the post-synaptic requirement for R7 connectivity. A complete understanding of the mechanisms of R7 target selection requires the identification of connectivity cues expressed by the R7 target neurons. To that end, we developed a genetic mosaic technique that allows manipulation of medulla target neurons as well as visualization of their axon projections. We are currently using this technique to test several candidate genes for their post-synaptic requirement in R7 target selection.
To manipulate medulla target neurons genetically, we needed to express yeast recombinase (flipase) before the neurons undergo the last cell division and commit to their final cell fates. First, we screened Gal4/lacZ enhancer trap lines from both public stock and our collections for expression in medulla precursor cells. We succeeded in identifying one such enhancer trap line that expresses Gal4 in both medulla and lamina precursor cells. Second, we constructed a flipase-based enhancer trap vector, P{flip}, and generated transgenic flies. Third, we used the P-element-swap technique to replace the exiting enhancer trap vector with our P{flip} vector. The resulting flies express flipase in medulla and lamina precursor cells during development. Given that the flip expression is under the direct control of the promoter identified in the original enhancer trap, we can use the flip system in combination with the existing MARCM (mosaic analysis with a repressible cell marker) system. We demonstrated that the P{flip}/MARCM system is capable of generating a large number of mosaic medulla neurons and a few mosaic lamina neurons. We are currently testing several candidate genes for their requirement in medulla target neurons for R7 connectivity.
In addition to directed mosaic analysis, the P{flip} enhancer trap system has two other applications: lineage tracing and combinatorial gene expression. To trace the lineage of the medulla precursor cells, we combined the P{flip} system with a conditional reporter, Actin<interruption cassette>Gal4 UAS-GFP. The flipase expressed in the medulla precursor cells removes the interruption cassette, thereby allowing the constitutive actin promoter to drive GFP expression in the progenitors of the medulla precursor cells even after flipase expression has subsided. Using this lineage-tracing system, we are able to trace the progenitors of the medulla precursor cells from late larval to adult stages. For combinatorial gene expression, we combined the P{flip} system with a Gal4 enhancer trap and a conditional reporter, UAS<interruption cassette>GFP. We were able to restrict the GFP (or any transgene) expression in the subset of neurons that expresses both flipase and Gal4. The combinatorial gene expression system has many potential uses for manipulating and analyzing small subsets of neurons in behavioral or histological studies.
Tracing color-vision circuitry
Ting, Gao; in collaboration with Meinertzhagen, Wang
To understand how color information is processed in the fly brain, we determined the connection patterns of the second-/third-order interneurons, which synapse with R7 (UV channel), R8 (blue channel), or lamina neurons (green channel). By screening Gal4 enhancer trap lines and subsequently analyzing them with the single-cell mosaic technique, we identified seven types of medulla interneuron neurons that likely receive input from one or two color channels. In particular, the small-field Tm5-type neurons extend dendritic arbors in the M2 (L2) and M3 (R8) layers and might convey both green and blue color information to the higher visual center lobula. Similarly, the small-field Tm20-type neurons connect both R7 (UV channel) and R8 (blue) inputs to a specific layer in the lobula. The Tm5 and Tm20 neurons might function as color-opposing neurons that calculate intensity differences between different spectra. These findings suggest that color-vision circuits in insects and primates share surprising similarity. Furthermore, our results highlight the lobula ganglion as the higher visual center for color vision. We are currently analyzing the connection patterns of these neurons at the electron microscopy level and testing how electrically silencing the neurons affects wavelength-selection behaviors.
COLLABORATORS
Andrew Chess, PhD, Whitehead Institute, Cambridge, MA
Akira Chiba, PhD, University of Illinois, Urbana, IL
Shu-ning Hsu, BA, University of Illinois, Urbana, IL
Ian Meinertzhagen, PhD, DSc, Dalhousie University, Halifax, Canada
Guilherme Neves, PhD, Whitehead Institute, Cambridge, MA
Jing Wang, PhD, University of California San Diego, San Diego, CA
For further information, contact leechih@mail.nih.gov.