Alteration of Drp1 Expression in Drosophila Models of Parkinson Disease

Background: Parkinson Disease (PD) and other neurodegenerative diseases have a signicant relationship with mitochondrial dysfunction. The substantial effect of mitochondrial dynamics in PD has led us to study the role of the gene encoding the mitochondrial ssion protein Drp1 in Drosophila models. Drp1 is a member of the highly conserved, dynamin family of protein encoding genes. Drp1 plays an essential role in the maintenance of the mitochondrial, peroxisomal, and endoplasmic reticulum (ER) dynamics, and has been found to regulate processes during homeostasis and cell survival. Results: The directed expression of Drp1 in Drosophila melanogaster neurons under the control of the Ddc-Gal4 transgene decreases the lifespan and compromises climbing ability over time. The directed inhibition of Drp1 produces a novel model of Parkinson Disease, as this causes little change in mean lifespan but a signicant decrease in locomotor or climbing abilities. Interestingly, the loss of park dependent Drosophila model of PD is rescued by the directed inhibition of Drp1. Conclusion: The compromised climbing abilities in ies with directed inhibition of Drp1 has produced a new model for Parkinson Disease and can be used to further investigate the mechanism(s) underlying PD and other neurodegenerative diseases. Interestingly, the combined inhibition of both Drp1 and park suppresses Parkinson Disease phenotypes and promotes survival.


Background
Mitochondria are critical organelles in the process of survival at the cellular level. Mitochondria accumulate damage over time to become dysfunctional and contribute to the process of ageing and eventually to the death of an organism [1]. Mitochondria are known to be the powerhouse of the cell, but its role goes beyond the production of ATP (oxidative phosphorylation); it is responsible for various aspects of energy homeostasis, oxidative stress, calcium handling, cell signalling, and cell survival [2].
The central role of mitochondria in energy regulation and signalling implies that its dysfunction would have devastating effects upon cellular functions. Such is especially true for the nervous system because subtle changes in signalling can have catastrophic consequences leading to neurodegenerative disease.
The shape and size of mitochondria are not xed properties; mitochondrial morphology depends upon a number of factors, including stage of the cell cycle and cell type, and can change quite quickly in response to external stimuli or metabolic cues. Changes to the mitochondrial network seem to differentially in uence a number of signalling pathways [3]. Therefore, mitochondria undergo ssion and fusion frequently in order to change their structure in response to the speci c requirements of the cell, under a wide range of circumstances. Mitochondrial ssion allows segregation of damaged mitochondria components while the process of fusion facilitates the exchange of mitochondrial material vital to maintain homeostasis within the mitochondrial network. Mitochondrial fusion helps compromised mitochondria, with highly damaged DNA and protein, to actively exchange components with other more healthy mitochondria to decrease the severity of heteroplasmy, and help with functional complementation [4,5]. Mitochondrial ssion allows for the segregation of irreversibly damaged portions of the mitochondrial network and subsequent degradation. Mitochondrial ssion necessarily requires dynamin-related protein 1 (Drp1) and FIS1 [6]. Nevertheless, an explicit understanding of the factors promoting ssion and fusion remains limited.
The Drp1 gene encodes a dynamin family GTPase protein comprised of a characteristic Dynamin like protein family domain (Dynamin and Mx protein domains), a dynamin central domain, and a dynamin GTPase effector domain. While predominantly located in the cytoplasm, a small fraction is located upon the cytoplasmic surface of the mitochondrial tubules. Overexpression of Drp1 causes mitochondrial fragmentation, whereas inhibition results in the elongation of the mitochondrial network. Drp1 protein function is regulated by post-translational modi cation via phosphorylation, where a well-documented phosphorylation site, S616, promotes ssion through an increase in activity. In contrast, phosphorylation of another site, S637, acts to lessen ssion through reduced activity [7,8,9]. Drp1 polymerizes to form a spiral structure around the mitochondrial tubule, and then utilizes its GTPase activity to constrict the tubule and eventually cause fragmentation of the mitochondria [10, 11, 12 13]. Presumably, a low membrane potential promotes ssion, while a high membrane potential obstructs ssion [14]. Through its role in mitochondrial ssion, Drp1 controls mitochondrial morphology and function.
The role of Drp1 protein is not limited to mitochondrial ssion. In addition, it participates in peroxisomal fragmentation [15] and maintains the morphology and function of the Endoplasmic Reticulum (ER) [16,17]. Also, Drp1 is required for a standard rate of cytochrome c release and caspase activation during programmed cell death [18]. The role of Drp1 in apoptosis is not clear, but the product of the Bax gene, a pro-apoptotic Bcl-2 family protein, has been found to co-localize with Drp1 at the site of mitochondrial ssion [19]. Drp1 plays role in Bcl-2 regulated apoptosis, peroxisomal and ER fragmentation.
The Drp1 protein interacts with other proteins involved in a number of mitochondrial processes, such as the products of Pink1 (PTEN-induced putative kinase 1) and park [20]. Mutation of the Pink1 and park genes are among the most prominent causes of early onset PD. The roles of Pink1 and park are vital to ubiquitin-dependent mitophagy. The protein product of park is involved in the process of ubiquitinylation that plays a crucial role in the proteasomal directed degradation of proteins, such as Drp1 [21]. Activated Akt can increase Drp1 phosphorylation and its localization to the mitochondria, to promote "mito ssion" with an accompanying increase in the generation of reactive oxygen species [22]. Drp1 knockouts increase nuclear translocation of transcription factor foxo and enhance the expression of its downstream targets [23]. Akt protein inhibits the nuclear translocation of foxo [24,25] and regulates Pink1/park dependent mitophagy [26]. Foxo transcriptionally upregulates the Pink1/park pathway in mammals and Pink1 appears to function upstream of foxo in D. melanogaster [27,28]. Such a feedback mechanism implies a complex relationship between Drp1 and park, Akt1 and foxo.
Here, we propose that gain-of-function of Drp1 results in a Parkinsonian-like phenotype. We use drosophila to model Parkinson disease because it is an excellent model system in which to study interactions between genes, including PD genes [29,30]. The inhibition of Drp1 in the dopaminergic neurons of a mouse MPTP model of PD gives protection against mitochondrial translocation of p53 and the loss of dopaminergic neurons [31]. In our experiment, we used the UAS-Gal4 system to direct and to inhibit the expression of the Drp1 gene in Drosophila. The GMR-GAL4 transgene directs expression in the developing eye [32] and the Ddc-Gal4 transgene directs expression in dopaminergic and serotonergic neurons [33,34]. Overexpression of Drp1 has toxic effects; although, its inhibition slightly improves the lifespan, the climbing ability over time is compromised. In an established park-RNAi model of PD [35], we directed and inhibited the expression of the Drp1 gene. The PD -like phenotypes of Ddc-Gal4 park RNAi were rescued by the expression of Drp1-RNAi transgenes. The strategy is to identify the basic mechanism in simple model organism and then further validate the nding in mammalian model organism.

Results
Drp1 is highly conserved between Homo sapiens and Drosophila melanogaster The D. melanogaster Drp1 protein sequence was sourced from NCBI protein and the conserved sequences were identi ed using NCBI CDD. NCBI protein Blast of Drp1 protein of D. melanogaster (NP_608694.2) with the H. sapiens, identi ed dynamin-1-like protein (isoform 4) (NP_001265392.1), it is 65% identical with a bit score of 957. The multiple sequence alignment of the two proteins derived by Clustal Omega (Fig. 1A) shows a highly conserved dynamin-like protein family domain, a dynamin central domain, and a dynamin GTPase effector domain. Two well-documented phosphorylation sites are identi ed; S606 and S627 in dynamin-1-like protein isoform 4 of H. sapiens; and S616 and T637 in Drp1 of D. melanogaster. A template-based modeling of D. melanogaster Drp1 protein by use of a combination of empirically derived energy functions and physics-based simulated folding was produced using Phyre2. The modeled D. melanogaster Drp1 protein (i) and the H. sapiens Dynamin-1 like protein (ii) from the NCBI database share an identical structure (Fig. 1B). The amino-terminus region of the Drp1 protein is highly conserved and has a consensus LC3-interacting region (LIR) sequence for binding to the ATG8/LC3 protein as determined by the Eukaryotic Linear Motif (ELM) resource. As this protein is so highly conserved, it seems very likely that the functions are highly conserved.

The overexpression and inhibition of Drp1 with Ddc-Gal4
The overexpression of Drp1 by the Ddc-Gal4 transgene results in a decreased lifespan of 56 days compared to the control of 68 days shown in Fig. 2A. Inhibition of Drp1 by two distinct RNAi transgenes, via the UAS-Drp1-RNAi1 and UAS-Drp1-RNAi2 directed by the Ddc-Gal4 transgene, results in lifespans of 70 and 72 days, respectively; very similar to the 68 days observed in the control, as determined by logrank at a P < 0.0001 ( Fig. 2A). While, the overexpression of Drp1 by Ddc-Gal4 does little to alter locomotion over time, the climbing abilities of ies expressing Drp1-RNAi's are severely compromised as determined by 95% con dence interval in a nonlinear tting of the climbing curve (Fig. 2B).
The inhibition of Drp1 in the Ddc-Gal4 UAS-park-RNAi model of PD The loss of function of park has led to the establishment of a number of Drosophila models of PD. The Ddc-Gal4 park-RNAi UAS-lacZ critical males have a median lifespan of 58 days. Overexpression of Drp1 in the Ddc-Gal4 park-RNAi expressing ies results in a similar life span of 58 days which is not signi cantly different compared to the control. The two UAS-Drp1-RNAi transgenes, UAS-Drp1-RNAi1 and UAS-Drp1-RNAi2, when expressed along with Ddc-Gal4 park-RNAi, results in a much-increased median life span of 84 and 76 days, respectively; determined by log-rank at a P < 0.0001 (Fig. 3A). The overexpression of Drp1 by Ddc-Gal4 along with park-RNAi slightly increases the climbing ability over time. However, the locomotor activity of the critical classes with the directed expression of the Drp1-RNAi transgenes are signi cantly increased (Fig. 3B) as determined by 95% con dence interval in a nonlinear tting of the climbing curve.
Overexpression of Drp1 during development of the eye decreases ommatidia and bristle number The inhibition and overexpression of Drp1, directed by the GMR-Gal4 transgene in the developing eye of ies affects development. The expression of UAS-Drp1 and UAS-Drp1-RNAi1 in developing eye directed by GMR-Gal4 transgene results in signi cantly higher mean number of ommatidia, 716.9 and 718.8, respectively compared to 703 for the lacZ control ies (Fig. 4B) as determined by unpaired T test with a P value of 0.0483 and 0.0484. The mean of interommatidial bristle produced through inhibition by the UAS-Drp-RNAi1 and UAS-Drp1-RNAi2 transgene was signi cantly higher at and 556.7 (P value = 0.0406) and 578.7 (P value = 0.0023) compare to 536 of control ies as determined by an unpaired T test. The mean number of interommatidial bristle for UAS-Drp1 ies was 541.6 compared to 536 of control, which is not signi cant as determined by an unpaired T test (P value = 0.6128). The ommatidia area, compared to the control of 200 um 2 per ommatidium, of the overexpression critical class was 213.7 um 2 (P value = 0.0303) and the ommatidium area produced by the UAS-Drp1-RNAi1 and UAS-Drp1-RNAi2 transgenes were 225 um 2 (0.0490) and 217 um 2 (P value = 0.0011), which are all signi cantly different as determined by unpaired T test (Fig. 4D).

Discussion
The critical role played by the structure of the mitochondria in the function of the organelle suggests that the product of the Drp1 gene acts as an essential component in the regulation of a number of sub-cellular processes. Excessive mitochondrial fragmentation is associated with dysfunctional metabolic diseases and a "hyper-fused" mitochondrial network serves to protect from metabolic insult and autophagy [36]. In the skeletal muscle of mice, Drp1 overexpression causes the severe impairment of post-natal muscle growth as it results in attenuated protein syntheses and the downregulation of growth hormone pathways [37]. The high fat and high glucose conditions cause excessive oxidative stress and mitochondrial fragmentation mediated by the Drp1 protein [38,39]. These phenotypes are similar to the overexpression of DLP1/Drp1 and inhibition of the MFN2 gene. In contrast, the levels of MFN2 mRNA are increased by acute weight loss [40]. The effect of Drp1 overexpression upon physiological processes is toxic.
The balance between mitochondrial ssion and fusion is very delicate. For example, a newborn girl with microcephaly, along with other developmental defects, was shown to have defective mitochondrial and peroxisomal ssion likely due to a dominant negative mutation in the Drp1 homologue [41]. Drp1 is essential for embryonic development in mice such that homozygous mutants or knockout Drp1 mice die during embryogenesis [42]. The Drp1 protein assists in caspase-independent mitochondrial ssion to amplify apoptosis [43]. The acute overexpression and inhibition of the Drp1 gene has shown similar phenotypes [23], most likely due to an extreme disruption in mitochondrial morphology.
Increasing the expression of Drp1 in the neurons of Drosophila under the Ddc-Gal4 transgene seems to decrease longevity, and Drp1 has been found to increase the ROS levels in mitochondria [44]. In our experiments, the lifespan of ies that overexpress Drp1 was signi cantly decreased when compared to controls, which may be due to elevated oxidative stress beyond a threshold of a normal lifespan. The inhibition of Drp1 in dopaminergic neurons through the directed expression of the Drp1-RNAi's results in a slight increase in lifespan. Drp1 overexpression and inhibition does signi cantly in uence the number of ommatidia or interommatidial bristles, and signi cantly increase ommatidia area when expressed under the control of the GMR-Gal4 transgene.
The well-documented function of Drp1 is the promotion of mitochondrial ssion and the inhibition of mitochondrial fusion [45]. Mitochondria host a number of cellular processes, especially oxidative phosphorylation, and are thus under continuous cellular stress and require repair and replacement [46,47,48]. The lower tolerance of mitochondria to ssion, compared to fusion, suggests that a continuous network of mitochondria can survive greater injury due to a lower rate of mitophagy and a slower rate of mitochondrial biogenesis, up-to an optimum level. The overexpression of Drp1 results in the excessive fragmentation of the mitochondria such that its e ciency can be diminished to the point that functional complementation by fellow mitochondria is negligible. Mitochondria can become a distinct burden for the cell to upkeep instead of being an e cient "powerhouse of the cell". Clearly, the level of Drp1 expression plays a key role in cell survival.
The park gene is crucial to the function of Pink1-dependent mitochondrial mitophagy. The loss of the park protein is a cause of great cellular stress as a major mitophagy mechanism is compromised to potentially result in the accumulation of non-functional mitochondria. Under normal circumstances, the park ubiquitin ligase recruits the Drp1 protein to mediate mitochondrial fragmentation during mitophagy [20,49]. In mouse embryonic broblasts, loss of park does not have any effect upon the number of mitochondria [50]. Furthermore, the loss of park along with the loss of Drp1 increases the number of mitochondria by threefold, which can be interpreted that park controls mitochondria fragmentation in a Drp1 knockout background. Therefore, park can negatively regulate Drp1-independent mitochondrial division. The impact of the loss of park can be observed only when the majority of mitochondrial ssion activity, mediated by Drp1, is absent.
In our experiments, the critical class ies that have the directed co-expression of Drp1-RNAi and park-RNAi inhibitory transgenes live longer than those that express park-RNAi and over-express Drp1 under the Ddc-Gal4 transgene. The inhibition of Drp1 and park though directed RNA interference promotes survival. This could be due to the establishment of an altered mitochondrial network to enhance homeostasis and bene t cellular health.

Conclusion
The Drp1 is not a major Parkinson disease gene, but it has been associated with cell death pathways in neurons. The overexpression of Drp1 gene in neurons results in the reduced survival and an agedependent decline in locomotor ability. The knockdown of Drp1 in the Ddc-Gal4-transgenes of Drosophila results in an age-dependent loss in climbing ability, phenotypes that are strongly associated with neuronal degeneration and Parkinson disease. Thus, the compromised climbing abilities in ies with directed inhibition of Drp1 have produced a new model of Parkinson Disease and can be used to investigate further the mechanisms underlying PD and other neurodegenerative diseases. The coinhibition of the parkin with Drp1 results in the rescue of the phenotypes observed, it is possible that Drp1 and parkin participate in cellular pathways that promote cell death. Further studies are required to better understand the interaction between parkin and Drp1 in these neurons. Overall our experiments are allowing us to contribute to our understanding of mitochondrial health and enhanced conditions of homeostasis.

Survival assay
Female virgins of the Ddc-Gal4 genotype were collected every 8 to 12 hours for several days. The con rmed virgin ies were then crossed with UAS-lacZ, UAS-Drp1, UAS-Drp1-RNAi1 and UAS-Drp1-RNAi2 males. Critical class male progeny were collected from mating's until approximately 250 ies of each genotype were obtained. To avoid over-crowding, the ies were maintained in cohorts of 25 or less per vial on standard media. Flies were scored every second day for viability and were transferred to new food every two to ve days. Scoring continued until all ies had died [55]). Longevity data were analyzed using GraphPad Prism version 8 statistical software (graphpad.com), and survival curves were compared by Mantel-Cox test. Signi cance were determined at 95% con dence level (P ≤ 0.05) with Bonferroni correction.

Locomotor analysis
Approximately 70 male ies of the critical class were collected within a 24-hour period and maintained as cohorts of 10 ies in each vial. Media was replenished twice a week. The climbing assay was performed as previously described according to a standard protocol [54,55]. Brie y, every week 50 males were assayed, in groups of 10, for their ability to climb a glass tube divided into 5 levels of 2 cm each. The climbing index was calculated for each week using GraphPad prism version 8 statistical software. The climbing curve was tted using non-linear regression and determined at a 95% con dence interval (P ≤ 0.05).

Biometric analysis of the Drosophila melanogaster eye
Female virgins of the GMR-Gal4 genotype were collected every 8 to 12 hours for several days. The con rmed virgins were then crossed with the males of UAS-lacZ, UAS-Drp1, UAS-Drp1-RNAi1 and UAS-Drp1-RNAi2 genotypes. Critical class male progeny were collected for each genotype. The collected ies were kept as cohorts of 10 ies or less upon fresh media and allowed to age for 3 to 4 days. The ies were prepared for scanning electron microscopy following the standard protocol [35]. Ommatidia and interommatidial bristle counts were performed on 10 or more ies of each genotype using National Institute of Health (NIH) ImageJ software. The ommatidium area was calculated by measuring the area of 5 distinct ommatidial "rosettes" per y eye and then dividing by 7 to determine the mean area of each ommatidium; done on 10 eyes of each genotype. The Biometric analysis was performed using GraphPad Prism version 8 statistical software. Signi cance were determined at 95% con dence level (P ≤ 0.05). Availability of data and materials All data generated or analyzed during this study are included in the article.

Competing interests
The authors declare no competing interests.  2) from a 76% identical protein with a con dence of 100%. The N terminus is coloured in Magenta; C terminus is coloured in Red and a consensus ATG8 binding region at N terminus is coloured in orange.  RNAi ies has signi cantly increased compared to control as determined in nonlinear tting of the climbing curve by 95% con dence interval.