Neural specific knockdown of cytochrome c oxidase and rescue of deleterious phenotypes with alternative oxidase in Drosophila melanogaster

*Background and aims:* Defects in mitochondrial respiratory chain produce diseases commonly referred as mitochondrial diseases. Mitochondrial diseases most often affect tissues that have high energy demands, particularly muscle and nervous system, and they are characterized by a plethora of symptoms. In this study, pathological effects of mitochondrial diseases were simulated in Drosophila melanogaster by neural specific cytochrome c oxidase knockdown. The aims of this research were to characterize the phenotype of neural specific COX knockdown in flies and test whether the alternative oxidase (AOX) can rescue the deleterious phenotypes. *Materials and methods:* Two UAS-RNAi constructs were used to knockdown COX subunits IV and Vb. RNAi was expressed in neurons by three neuronal GAL4 drivers. To verify COX knockdown in nervous system, sections of paraffin embedded fly muscle and brain were stained with fluorescent antibodies. The expression of AOX was also confirmed using the same method. The phenotype of neuronal COX-KD flies was characterized by i) viability of the flies (lethality assay) and ii) degenerative loss of viability after eclosion (survival assay). The same experiments were conducted on flies that were coexpressing AOX or Ndi1, along with COX knockdown. Additionally, the expression pattern of two elav-GAL4 drivers was determined with GFP in different developmental stages in different tissues. *Results:* Neural specific COX-KD using elav-GAL4 drivers presented variable phenotypes, including partial and full lethality, locomotor defects and reduction in viability. The phenotypes were stronger in COX IV-KD flies compared to COX Vb-KD flies and also in males compared to females. AOX expression was able to partially or fully rescue the deleterious phenotypes due to COX-KD. Surprisingly, Ndi1 expression made the phenotypes even worse. AOX was found to be expressed in the thoracic muscles of flies. Closer inspection of the expression pattern of elav-GAL4 drivers revealed that both of them drive the expression of UAS transgenes also in thoracic muscles at pupal and adult stages. *Conclusions:* Dosage compensation of X-chromosomal genes can explain the stronger phenotype in male flies. This is true particularly in the case of elav(X)-GAL4. The stronger phenotype in COX IV-KD flies reflects the role of subunit IV in early stages of COX biogenesis. Therefore COX IVKD is expected to lead to more severe COX dysfunction than COX Vb-KD, which is used in later stage of COX synthesis. AOX was able to fully or partially rescue deleterious phenotypes, which is attributable to its ability to restore proton flow through complex I of the respiratory chain. Thus, ATP production is restored to a level that allows vital cellular functions to operate. Ndi1 expression seems to have the opposite effect. Therefore the level of ATP is expected to be further decreased. This and previous studies show that AOX has true potential to be a therapeutic agent for OXPHOS dysfunction