Physiological roles of Drosophila ADAR and modifiers

ADAR (Adenosine Deaminases acting on RNA) family proteins are double-strand RNA binding proteins that deaminate specific adenosines into inosines. This A-to-I conversion is called A-to-I RNA editing and is well conserved in the animal kingdom from nematodes to humans. RNA editing is a pre-splicing event on nascent RNA that may affect alternative splicing when the editing occurs in the exon-intron junction or in the intron. Also, editing may change biological function of small RNAs by editing the premicroRNAs or other noncoding RNAs. Editing also alters protein amino acid sequences because inosine in the mRNA base pairs with cytosine and is therefore read as guanosine. In mammals, there are three ADAR family proteins, ADAR1, ADAR2, and ADAR3, encoded by three different genes. So far, no enzymatic activity of ADAR3 is detected. The most frequently edited targets of ADAR1 and ADAR2 are regions covering copies of Alu transposable elements in primates. In addition, loss of some specific editing events leads to profound phenotypes when the editing does not occur correctly. For example, some human neural disorders – such as epilepsy, forebrain ischemia, and Amyotrophic Lateral Sclerosis – are known to be associated with abnormally edited ion channel transcripts. Drosophila has a single ADAR protein (encoded by the Adar gene) that is highly conserved with human ADAR2 (encoded by the ADARB1 gene). To date, 972 editing sites have been identified in 597 transcripts in Drosophila, and approximately 20% of AGO2-associated esiRNAs are edited. Similar to mammals, many ion channel-encoding mRNA transcripts undergo ADAR-mediated A-to-I editing in Drosophila. While Adar1 null mice die at the embryonic stage and Adar2 null mice die shortly after birth due to seizures, Adar null flies are morphologically normal and have normal life span under ideal conditions. However, Adar null flies exhibit severe neurodegeneration and locomotion defects from eclosion, whilst Adar overexpression (OE) is lethal. To better understand the physiological role of RNA editing and ADAR, and to shed light on ADAR-related human disease, I used Drosophila Adar mutant flies as a model organism to investigate phenotypes, and to find chromosomal deletions and specific mutations that rescue the neural-behavioural phenotype of the Adar null mutant flies. Using the publicly available chromosomal deletions collectively covering more than 80% of the euchromatic genome of Chromsome III, I performed a genetic screen to find rescuers of the lethality caused by Adar overexpression. I confirmed that mutation in Rdl (Resistant to dieldrin, the gene encoding GABAA receptor main subunit) rescues. This rescue was not likely caused by effects on Adar expression level or activity. Driven by the hypothesis that the rescue may be due to reduction in GABAergic input to neurons, I recorded spontaneous firing activity of Drosophila larval aCC motor neurons using in vivo extracellular current recording technique. As expected, the neurons overexpressing Adar had much less activities compared with wild type neurons. Also, I found that Adar null fly neurons fired much more and showed epilepsy-like increased excitability. Although feeding PTX (Picrotoxin), a GABAA receptor antagonist, failed to rescue the lethality, reducing the expression of GAD1 to reduce synthesis of GABA was able to rescue the ADAR overexpression lethality. These results suggest that ADAR may finetune neuron activity synergistically with the GABAergic inhibitory signal pathway. I used MARCM (mosaic analysis using a repressible cell marker) to detect cellautonomous phenotypes in Adar null cells in otherwise wild type flies. Although neurodegeneration, observed as enlarged vacuoles formation in neurophils, was detected both in histological staining and EM images, the Adar null neurons marked with GFP from early developmental stages were not lost with age. Nevertheless, swelling in the axons or fragmentation of the axon branches of Adar null neurons was sometimes observed in the midbrain. By comparing the Poly-A RNA sequencing data from Adar null and wild type fly heads, we detected significant upregulation of innate immune genes. I confirmed this by qRT PCR and found that inactive ADAR reduces the innate immune gene transcript levels almost as much as active ADAR does. Further, using the locomotion assay, I confirmed that reintroducing inactive ADAR into Adar null flies can improve the flies’ climbing ability. Based on the Adar null flies having comparatively low viability, I performed a second deficiency screen to find rescuers of Adar null low viability using the same set of deficiencies as in the lethality rescue screen described above. I found seven deletions removing 1 to 37 genes that significantly increased the relative viability of the Adar null flies. However, not all the rescuing deficiencies also improved the Adar null locomotion. One rescuing gene, CG11357 was mapped from one of the rescuing deficiencies, and some mutant alleles of cry, JIL-1 and Gem3 also showed significant effects on the Adar null fly viability. The single gene viability rescuers were also not necessarily locomotion or neurodegeneration rescuers. Although the initial aim was to find neural-behavioural rescuing genes from the viability screen, the viability rescuers found in the screen are more likely to play a role in different aspects of stress response for survival