Dscam2, a cell surface protein that mediates cellular repulsion, plays a crucial role in the development of the Drosophila melanogaster visual system. Dscam2 generates boundaries between neighboring modules in the fly optic lobe; in Dscam2 mutants this visual system modularity is compromised. Although developmental wiring defects have been well described in the Dscam2 mutant, behavioral consequences have not been investigated. To address this, we examined the visual behavior of Dscam2 mutant flies. Using a phototaxis assay, we ascertained that these flies are not blind, but have a reduced phototaxic response. Through population-based and single fly optomotor assays, we found that Dscam2 mutant flies can track motion but that their response is opposite to control flies under defined experimental conditions. In a fixation paradigm, which allows tethered flies to control the angular position of a visual stimulus, mutant flies' responses were diametrically opposed to those seen in control flies. These data suggest that modest changes in the modularity of the fly visual system in the Dscam2 mutant can dramatically change the perception of specific visual cues and modify behavior.

The patterns of synaptic connections between neurons in the brain are key determinants of behaviour, but how specific neural circuits lead to distinct behaviours is largely unknown. Cell recognition molecules that mediate interactions between neurons play a crucial role in establishing synaptic connections. Down Syndrome Cell Adhesion Molecule 2 (Dscam2) is required for establishing modularity in the visual system. Dscam2 null flies have a disorganized visual system, lack boundaries between neighbouring modules in the optic lobe and exhibit changes in the postsynaptic composition of photoreceptor synapses. The behavioural consequences of these wiring defects have not been explored previously. In this thesis, I begin to dissect how these changes in neural circuitry affect visual system behaviours such as the optomotor response, object-orientation preference, and attention-like object-tracking. In order to do so, a population and single tethered-fly assay have been optimized and used to explore motion perception by assessing the visual response to a wide range of psychophysical parameters of black/green moving gratings. In addition, a single tethered-fly virtual reality assay has been set-up and used to explore orientation preference and attention-like tracking.Through the population and single tethered-fly optomotor assays, it is shown that Dscam2 null flies can track motion but that their response is opposite to control flies under defined experimental conditions.Through the single tethered-fly virtual reality assay, it is shown that Dscam2 null flies anti-fixate on a dark bar, again the opposite behaviour to control flies. Responses to a light bar are in contrast, the same as controls. It is shown that this anti-fixation is bar width dependent. Together these results demonstrate that the disrupted visual system of the Dscam2 null flies can dramatically change the perception of specific visual cues and modify behaviour. Lastly, other perturbations of Dscam2 were studied such as Dscam2 single isoform and Dscam2 trisomic flies. Dscam2B single isoform flies displayed some motion detection phenotypes. In addition, they were not responding to either a dark or light bar in the tethered virtual reality assay. Dscam2 trisomic flies were able to detect gratings with different spatial and temporal frequencies but had a change in orientation behaviour as they displayed fixation behaviour to both a dark and a light bar. The results provide a foundation for understanding how brain miswiring can lead to changes in behaviour.iii In chapter 1, I outline the background information to provide context for the experiments and overall research aim. In chapter 2, I contrast a population and single fly assay to explore motion detection in Drosophila.In chapter 3, I present a new single fly virtual reality assay which goes beyond simple visual reflexes by measuring object-orientation preferences and attention-like tracking. In chapter 4, I test motion detection in Dscam2 null flies. In chapter 5, I test object-orientation prefe...