Mitochondria are “the powerhouses of the cell” playing a crucial role in the control of intracellular metabolism. The main processes in mitochondria are electron transport and proton pumping with the energetic steps (oxidative phosphorylation) that harness the energy as ATP. However, these processes are not without risk for cells, and the electron transport chain is sensitive to multiple external stressors. One of the functional disturbances within mitochondria is the stressed-induced massive production of reactive oxygen radicals e.g., the superoxide anion (O₂^•-). These dismutate toward hydrogen peroxide (H₂O₂), which can subsequently react to become hydroxyl radicals (HO^•) that are most harmful and destructive to cells (Sena and Chandel, 2012; Kehrer and Klotz, 2015). Numerous proteins, such as ligases and mitochondrial kinases, are responsible for the control of mitochondrial stability and their biological maintenance, with the main function of mitophagy (Youle and Narendra, 2011). This process involves the removal of damaged, old, or improperly functioning mitochondria by autophagy (Cornelissen et al., 2018). Disrupting mitochondrial processes may lead to the development of many diseases such as Parkinson’s disease (PD), a neurodegenerative disease. Given their high energy demands, neurons are cells that are most sensitive to mitochondrial damage (Trevisan et al., 2018; Zilocchi et al., 2018).

Two important and well-described mitochondrial proteins which control the stability of mitochondria are the Mul1 and Park E3 ubiquitin ligases. These proteins are responsible for promoting mitophagy and maintaining mitochondrial integrity and fusion-fission processes in Drosophila (Kitada et al., 1998; Yun et al., 2014). Mul1 is also involved in SUMOylation. Mutations in the genes encoding Mul1 and Park in Drosophila lead to typical PD symptoms such as motor disorders, sleep problems and degeneration of dopaminergic neurons (Clark et al., 2006; Park et al., 2006; Yun et al., 2014; Gokcal et al., 2017). The above symptoms may also be caused by various neurotoxins, one of which is rotenone. The mechanism of its action is based on the disruption of electron transport in mitochondria. It inhibits the transport of electrons from iron-sulfur centers in complex I on ubiquinone (Lindahl and Öberg, 1961). As a result, it triggers mitochondrial damage by increasing oxidative stress, leading to neuronal death. However, cells can counteract these changes by enhancing the activity of antioxidative enzymes i.e., catalase, superoxide dismutase, heme oxygenase-1, or glutathione peroxidase. All these proteins protect cells from oxidative stress-mediated programmed cell death, or apoptosis (Silva and Coutinho, 2010).

Neurodegenerative diseases can be studied using animal models, including the fruit fly Drosophila melanogaster. The Drosophila genome carries homologs of most of the genes involved in the development of Parkinson’s disease, with the notable exception of α-synuclein (Nagoshi, 2018). In addition, current genetic tools and their short period of development, allows successful manipulation of its genome to be performed (Duffy, 2002). Symptoms typical of Parkinson’s disease, e.g., dopaminergic neuron degeneration and motor disorders, can be induced in Drosophila by various neurotoxins, such as rotenone, which has been used in the present study, and MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine). Both toxins induce symptoms typical of Parkinson’s disease via mechanisms linked to oxidative stress (Coulom and Birman, 2004; Abolaji et al., 2018).

In the present study, we examined whether the strong expression of two major mitochondrial ligases may protect flies exposed to rotenone, against developing symptoms typical of Parkinson’s disease. We found that overexpressing genes encoding Mul1 and Park in all neurons in the Drosophila brain inhibits degeneration of dopaminergic neurons and the motor disorders caused by rotenone. In addition, we found that rotenone affects the structure of synapses and the expression of synaptic proteins in the brain of flies, but when the levels of Mul1 and Park were increased in parallel, synapse structure and the normal level of synaptic proteins were restored.

Rotenone is a toxin that inhibits activity of the mitochondrial complex I and increases production of the radical oxygen species level, as shown in rats (Betarbet et al., 2000). In the present study, we showed that exposure to 500 μM rotenone in Drosophila (Coulom and Birman, 2004) reduces the levels of Atg5 and Sod1, which are important for cell survival under stress. Sod1 is an enzyme that is responsible for the elimination of free radicals from cells (Bunton-Stasyshyn et al., 2015). Its low-level leads to hypergeneration of reactive oxygen species in the cell and its presence triggers cellular damage (Henchcliffe and Beal, 2008). Atg5 is one of the major proteins involved in autophagy, during which damaged organelles can be removed and amino acid obtained for cell survival and homeostasis in mice (Kuma et al., 2004). The enhanced autophagy during high oxidative stress is beneficial for cells and delays the degenerative processes in rat cells (Chen et al., 2014). Its disruption is critical for PD pathogenesis in non-dopaminergic neurons and for the onset of non-motor symptoms in rats (Wise et al., 2018). The suppression of autophagy has an adverse effect on the elimination of free radicals (Navarro-Yepes et al., 2014) and we demonstrated that rotenone also affects apoptosis by increasing the level of Dcp-1 in fruit flies (Xu et al., 2009). It has already been observed that rotenone exposure induces changes in the level of several apoptotic proteins such as Bcl2, Bax, Caspase-8, and Cyt-C in rats (Dhanalakshmi et al., 2016) and that autophagy prevents oxidative stress-dependent apoptosis (Li et al., 2017).

Our results also showed that rotenone exposure disturbs climbing ability and locomotor activity of flies, during the day. This is most likely caused by increased free radical levels and apoptosis which have been observed in rats (Rosiñczuk et al., 2018), particularly in neurons, leading to neuronal disorders. It has been reported that oxidative stress is associated with behavioral disorders in mice (Camargo et al., 2018) and we demonstrated that the mul1 or park overexpression in neurons can restore the normal level of apoptosis and increase autophagy and endogenous antioxidant enzyme levels. As a result, improvements in climbing and locomotor activity were observed. The results obtained also showed that overexpressing the ligases studied causes hyperactivity in flies and reduces their sleep time during the day. It has already been reported that increases in the level of mitophagy results in increased motor activity levels in fruit flies (Chakraborty et al., 2018) and this can explain hyperactivity in flies with higher levels of mitochondrial ligases. Mul1 and Park enhance mitochondrial fusion, so increasing their amount in neurons, most likely leads to the increased number of large mitochondria. In turn, large mitochondria produce more ATP than smaller ones (Sun et al., 2014), which in turn may increase the motor activity of Drosophila. Higher levels of Nix protein, a mitochondrial autophagy receptor, likewise increases ATP levels in strains that are also genetic models of Parkinson’s disease and shows that ATP levels depend on the quality of mitophagy (Koentjoro et al., 2017). Appropriate mitochondrial quality can also be provided by Afadin 6, an F-actin binding multidomain-containing scaffolding protein. This protein interacts with Park, as also reported in fruit flies, and its overexpression restores the physiological phenotype in the pink and park mutants (Basil et al., 2017). Our results suggest that the increased expression of park or mul1 can also lead to an increase in the level of Afadin 6, which in turn is protective for mitochondria. Results obtained by other authors also showed that the park mutation in Drosophila decreases mass and cell size and increases sensitivity to oxygen radical stress (Pesah et al., 2004). The overexpression of park studied here may also increase cell size and reduce the cell sensitivity to free radicals, which was also a protective effect against rotenone. Results from other authors have shown that overexpressing the gene encoding Lrrk2, a leucine-rich repeat kinase, also involved in the development of PD, inhibits degeneration of dopaminergic neurons in the Drosophila model using rotenone (Ng et al., 2009). This inhibition suggests that overexpressing genes, the mutation of which cause symptoms of Parkinson’s disease, might be protective against sporadic forms of the disease in animal models exposed to neurotoxins. The Usp30 protein also appears to be associated with the development of Parkinson’s symptoms. This protein is located in mitochondria and acts as an inhibitor of mitophagy. It has been shown that overexpressing this protein removes ubiquitin attached by Parkin to damaged mitochondria and blocks the ability of Parkin to drive mitophagy (Bingol et al., 2014). We suggest that Usp30 activity is suppressed by the increased levels of Park and Mul1, which restore the correct mitophagy.

Behavioral disabilities may also result from damaged synapses in the nervous system and neuromotor junctions after exposure to rotenone. We observed that rotenone decreases the level of several proteins associated with synaptic transmission: Dlg1, Synapsin, and Synaptotagmin. Dlg1 is responsible for clustering neurotransmitter receptors and ion channels in the postsynaptic membrane and for mediating cell-cell adhesion (Kim et al., 2014). Synapsin is important for vesicle clustering in the presynaptic site and this protein also regulates synaptic plasticity (Vasin et al., 2014). In turn, Synaptotagmin acts as a Ca²⁺ sensor for fast neurotransmitter release (Geppert et al., 1994). Low levels of Synapsin or Synaptotagmin lead to poor synaptic transmission and motor disorders in rats (Sampedro-Piquero et al., 2014; Lai et al., 2015). Moreover, a low level of synaptic proteins is also correlated with a high level of free radicals, which as described above, reduce the amount of Sod1, increases apoptosis and decreases autophagy. However, the low level of a synaptic protein also results from the weak activity of translational factors of this protein translation, because of the action of free radicals. Lee et al. (Lee et al., 2018) have reported that a high level of ROS in human cells affects the activity of eIF2α translation factor. As shown by our results, rotenone decreases the level of proteins, synaptic proteins, Atg5 and Sod1, while the expression of their encoding genes remains normal. In order to confirm the lack of synaptic proteins affects, we conducted a TEM study of tetrad synapses in the first neuropil of the optic lobe of D. melanogaster as a model population of synapses. We found that rotenone intoxication affects the morphology of synaptic vesicles as well as the synaptic active zone, which are responsible for neurotransmitter exocytosis into the synaptic cleft. The active zone within the presynaptic T-bar is a site where Ca²⁺ triggered fusion of a synaptic vesicle occurs (Wichmann and Sigrist, 2010). The proper development of synapses depends on autophagy in Drosophila (Shen and Ganetzky, 2009) which is disrupted by rotenone as well as being affected by oxidative stress (Milton and Sweeney, 2012; Kamat et al., 2016). The observed changes in synapse morphology were not observed in strains with mul1 and park overexpression. The T-bar active zones looked normal in both strains, T-bars were large with a large platform, and synaptic vesicles were round, dense in the middle, which is typical for well-functioning synapses.

Finally, we examined dopaminergic neurons after rotenone exposure in control and mul1 or park overexpression strains. The results of other authors have shown that rotenone at a concentration of 500 μM causes degeneration of these neurons in five clusters: PAL, PPL1, PPL2, PPM1/2, and PPM3 (Coulom and Birman, 2004). These authors did not describe the VUM cluster, but the images included in their article show no reduction of the number of neurons in this group. This was also confirmed in another report (Navarro et al., 2014). The decrease of dopaminergic neuron numbers is associated with the reduction of the dopamine level, which is probably associated with motor disability (Smith et al., 2013; Santiago et al., 2014). After rotenone exposure we observed a decreased number of dopaminergic neurons in the five clusters of neurons previously mentioned, while the overexpression of mitochondrial ligases in neurons of strains fed with rotenone was protective against the toxin and prevented degeneration of the dopamine neurons. Given the association between dopaminergic neurons and motor activity, we propose that the inhibition of their degeneration probably results in the improved climbing ability of Drosophila.

To conclude, in the present study we showed that rotenone at a concentration of 500 μM reduces the level of autophagy, apoptosis and increases the level of free radicals. This leads to the disrupted function and morphology of synapses, locomotor activity and dopaminergic neurons in the brain of D. melanogaster. We also demonstrated that the overexpression of two main mitochondrial ligases: Mul1 and Park in all neurons is protective, since they inhibit the effect of rotenone. The degeneration of neurons seems to depend not only on the lack of ATP, because of the damaged mitochondria, but also because of their destructive effect in the cell.

EP was responsible for funding acquisition, conceptualized the study, administered the project, provided the resources, supervised the study, and also wrote, reviewed and edited the manuscript. BD performed data curation and formal analysis and visualized the study. BD and MD investigated the study and performed the methodology. MD and EP validated the study. BD, MD, and EP wrote the original draft of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.