Institut für Biologie III, Albert-Ludwigs-Universität Freiburg
We have used the Gal4/UAS targeted expression system to analyze the regional specificity and developmental dynamics of reporter gene expression in a collection of enhancer trap lines expressing the yeast transcription factor Gal4 in subsets of cells within the optic lobe of Drosophila. As a reporter, we used a tau-GFP fusion construct. This fusion protein is transported into axonal processes due to the microtubuli-binding properties of the tau protein. It offers a higher level of resolution in structural analysis than either a conventional GFP construct or tau-lacZ fusions. Due to its natural fluorescence, it is ideally suited for combination with antibody stains for two-channel confocal laser scan microscopy.
We have used tau-GFP to analyze expression patterns in four Gal4 lines (Mz507, Mz1407, Mz1369, Mz 1525, all obtained from J. Urban and G.M. Technau, Mainz) that produce axonal projection phenotypes when crossed to UAS-irreC-rst. Two of these lines (Mz1525, MZ1407) show projection defects mostly restricted to the inner chiasm, while in the others (Mz507, Mz1369), both optic chiasms develop abnormally. Comparison of the phenotypes seen in UAS-irreC-rst transformants with the expression patterns revealed with UAS-tau-GFP allow to test experimentally models for the action of IrreC-rst in optic chiasm formation. Double staining experiments stress the fact that IrreC-rst marks young axon bundles during optic chiasm formation.
We have also analyzed adult serial sections and larval stages of tau-GFP-expressing animals to ensure that no developmental defects are induced by the fusion protein. We observed that, in contrast to earlier investigations on similar reporter fusions, the tau fragment seems to be biologically active in Drosophila and interferes with neuronal differentiation and axonal pathfinding. Two of four investigated lines showed defects that we judge to be tau-mediated as effects of tau-GFP are similar, though less severe, to those of a complete UAS-tau construct we used as a control. Mz507/UAS-tau-GFP animals were viable as adults but showed irregular agglomerations of axons in the lamina. In Mz1369, tau-GFP expression, as well as tau expression, proved lethal. In the optic lobe, photoreceptor differentiation was inhibited and the lamina did not develop. Animals did not develop beyond the early pupal stage.
These results show that, although tau fusion proteins offer an increased resolution of structural analysis in the developing nervous system, lines used have to be carefully checked for possible defects induced by the fusion protein and expression patterns should be closely compared to those seen with various reporter or effector constructs. However, studies using Drosophila expression systems may also offer novel insights into the functional capabilities of proteins such as tau.
