Drosophila lacking RNA editing
ADAR is an adenosine deaminase that acts on dsRNA. Once bound to dsRNA, ADAR deaminates specific adenosines into inosines. If this occurs within the coding region of a transcript the inosine will be read as a guanosine. This can lead to a change in the amino acid at this position and increase protein diversity. In mammals there are three ADAR genes: ADAR1, ADAR2 and ADAR3. However, only ADAR1 and ADAR2 have been shown to be enzymatically active. ADAR1 is widely expressed and can edit both coding RNA and non-coding RNA. ADAR2 is restricted to the CNS and the key transcript that it edits encodes the GluR-B subunit of the glutamate-gated ion channel receptor. Editing of the Q/R site in the GluR-B transcript occurs with an efficiency of more than 99.9% and changes the genomically encoded glutamine into an arginine. This results in an ion channel that is impermeable to calcium. The ADAR2 knock-out mice are viable, but suffer from epileptic seizures and die by day 20. This phenotype can be rescued by expressing the edited R isoform of GluR-B, suggesting that this site is the most important target for ADAR2. Drosophila has only one Adar gene and its product has been reported to edit more than one hundred adenosines in different transcripts. Many of these transcripts encode subunits of ion channels, and it has been hypothesised that lack of ion channel editing causes the behavioural defects and age-related neurodegeneration observed in Adar deletion mutants. In this thesis I investigate the function of ADAR in an uncharacterised Adar mutant, Adar5G1. To characterise the Adar5G1mutant I not only used standard histology but a 3D imaging technique, optical projection tomography (OPT), that had not been reported to be used with Drosophila before this work. OPT allows the internal organs to be imaged without any manual sectioning or dissecting. I used OPT to identify neurodegenerative vacuoles from within the intact head and present the data both in 2D and in 3D. In addition to this, I demonstrate that this technique can be used to image global expression patterns in the Drosophila adult and I relate the TAU-β galactosidase expression pattern to the Drosophila anatomy. The neurodegeneration observed by OPT was confirmed by detailed analysis of stained wax sections. Complete loss of Adar, in the Adar5G1 mutant revealed age-dependent vacuolisation of the retina and mushroom body calyces. The vacuolisation observed in the Adar5G1 mutant was rescued by expression of Drosophila Adar and human ADAR1 p110, and ADAR2. However the cytoplasmic form of ADAR1, ADAR1 p150, did not rescue the vacuolisation of the Adar5G1 mutant. ADAR3, a catalytically inactive ADAR, rescued the vacuolisation phenotype of the Adar5G1 mutant, suggesting that ADAR may have an additional function independent of editing activity. The vacuolisation of the Adar5G1mutant was found not to be associated with type I programmed cell death. However, it was associated with swollen nerve fibres and degrading ommatidia containing multilamellar whorls. Neurodegeneration in various Drosophila mutant models and human neuropathies has been associated with similar cellular structures, suggesting that loss of ADAR results in neurodegeneration common to many of the known neuropathies. Finally, I found that expression of edited isoforms of the nicotinic receptor channel 34E subunit (Nic 34E) failed to rescue the locomotion phenotype of the Adar mutant. However, I found preliminary evidence that one of the lines generated for an edited isoform of Rdl, a subunit of the GABA receptor ion channel, gave a partial rescue of both locomotion and neurodegeneration of the Adar1F4 and Adar5G1 mutant.