parkin counteracts symptoms in a Drosophila model of Parkinson's disease

Background Parkinson's disease, a prevalent neurodegenerative disease, is characterized by the reduction of dopaminergic neurons resulting in the loss of motor control, resting tremor, the formation of neuronal inclusions and ultimately premature death. Two inherited forms of PD have been linked to mutations in the α-synuclein and parkin genes. The parkin protein functions as an ubiquitin ligase targeting specific proteins for degradation. Expression of human α-synuclein in Drosophila neurons recapitulates the loss of motor control, the development of neuronal inclusions, degeneration of dopaminergic neurons and the ommatidial array to provide an excellent genetic model of PD. Results To investigate the role of parkin, we have generated transgenic Drosophila that conditionally express parkin under the control of the yeast UAS enhancer. While expression of parkin has little consequence, co-expression of parkin with α-synuclein in the dopaminergic neurons suppresses the α-synuclein-induced premature loss of climbing ability. In addition directed expression of parkin in the eye counteracts the α-synuclein-induced degeneration of the ommatidial array. These results show that parkin suppresses the PD-like symptoms observed in the α-synuclein-dependent Drosophila model of PD. Conclusion The highly conserved parkin E3 ubiquitin ligase can suppress the damaging effects of human α-synuclein. These results are consistent with a role for parkin in targeting α-synuclein to the proteasome. If this relationship is conserved in humans, this suggests that up-regulation of parkin should suppress α-synucleinopathic PD. The development of therapies that regulate parkin activity may be crucial in the treatment of PD.


Background
Parkinson's disease (PD) is a neurodegenerative disorder that is characterized by muscle tremors in stationary limbs, bradykinesia (slowed movement) and difficulty initiating and sustaining movements, and affects 1-2% of the population older than sixty years of age [1][2][3][4][5]. As the disease progresses, both the sense of balance and the memory of the affected individual deteriorate. Post-mor-tem analysis reveals the selective loss of dopaminergic neurons from the substantia nigra region of the brain. Filamentous protein inclusions, known as Lewy bodies, are found within the neuronal cell bodies of the affected area, in most but not all PD patients [4]. Although the majority of PD cases appear to be sporadic, about 5-15% have been determined to have an inherited basis [6,7]. Recently, mutations in a number of genes have been identified as causes of PD and many of these genes are associated with the ubiquitin/proteasome protein degradation pathway.
Autosomal Recessive Juvenile Parkinson's disease (ARJP), another inherited form of PD, has been attributed to a number of point mutations and deletions of the parkin gene [17][18][19]. ARJP is specifically characterized by a very early age of onset, mostly before forty years of age, and the absence of Lewy bodies [20][21][22]. In humans, the parkin gene encodes a 465 amino acid protein [17] that functions as one of a number of E3 ubiquitin protein ligases, components of the ubiquitin/proteasome degradation pathway [23]. Ubiquitin protein ligases act to identify damaged, misfolded, and short-lived proteins to mediate the ubiquitination (the sequential attachment of a number of ubiquitin monomers) of these proteins, which are targeted to the proteasome [24,25]. Experiments in tissue culture have demonstrated that parkin can ubiquitinate a number of substrates including a glycosylated form of α-synuclein [26], the Pael receptor [27], CDCrel-1 [28], the α-synuclein-binding protein synphilin-1 [29], and parkin itself [28,30]. The loss of parkin may lead to an accumulation of one or a number of proteins in sufficient quantities to cause neuronal cell death.
The interaction of parkin with α-synuclein suggests a common mechanism underlying inherited forms of PD. Indeed, elevated expression of parkin protects neuronal explants from the toxicity associated with expression of αsynuclein [26,31]. The disease inducing-forms of α-synuclein may prevent its degradation and result in toxic accumulation. In ARJP, functional parkin protein is lost along with the ability to mediate the ubiquitination of glycosylated α-synuclein and may lead to the accumulation of this protein [26]. Our working hypothesis is that aspects of parkin-mediated protein degradation are compromised in PD.
The first Drosophila melanogaster model of PD was generated by the conditional expression of human α-synuclein in transgenic Drosophila [32]. Flies that express α-synuclein, in either a pan-neural or dopaminergic neuron specific manner, show a marked age-dependent loss of dorsalmedial dopaminergic neurons. Cytoplasmic inclusions were observed in α-synuclein-expressing flies approximately 20 days after eclosion. While control flies exhibit a strong negative geotaxis, these transgenic flies prematurely lose their climbing ability. In addition, expression of α-synuclein in the developing eye results in precocious degeneration of the retina. In this model expression of a number of genes are dysregulated prior to the onset of neurodegeneration [33]. These features recapitulate the main behavioural and pathological phenotypes of PD and provide an excellent model system to study the biological basis of the disease.
The Drosophila α-synuclein based model has been used to investigate a number of aspects of PD. Pharmacological agents, such as the dopamine precursor levodopa, dopamine receptor agonists (bromocriptine, pergolide and SK&F38393), and the anticholinergic atropine, were demonstrated to modify the age-dependent loss of climbing ability [34]. Co-expression of the molecular chaperone HSP70 gene with α-synuclein prevented dopaminergic neuronal degeneration [35]. Interference with endogenous chaperone activity accelerated the toxicity of αsynuclein demonstrating a role for chaperones in the pathology of the disease. Suppression of HSP90, a negative regulator of the heat shock factor 1, by feeding flies geldanamycin prevents dopaminergic neuronal cell death [36]. Recently, the expression of human parkin has been shown to suppress the loss of dopaminergic neurons induced by α-synuclein in Drosophila [37]. This model has proven to be an effective tool in the investigation of the biological basis of PD.
To investigate the role of parkin in the α-synuclein-based model of Parkinson's disease in Drosophila, we have characterized and expressed the Drosophila homologue of parkin in this model. Our results demonstrate that parkin can counteract the effects of α-synuclein on climbing activity and retinal degeneration.

Characterization of Drosophila melanogaster parkin
The Drosophila melanogaster parkin homologue was identified through a search of the Berkeley Drosophila Genome Project (BDGP) utilizing the tBlastn search algorithm. The parkin gene is located on the left arm of the third chromosome, in the polytene chromosome section 78C within the genomic scaffolding region AE003593 (BDGP), and consists of 6 exons over 2.2 kb ( Figure 1A). A search of the genome for additional parkin homologues revealed none. Our analysis confirmed the sequence of parkin to be identical to that reported by the BDGP (AY058754.1). Two potential initiation codons for the parkin protein are separated by 42 base pairs at the 5' region of the transcript. As the second potential start codon is preceded by CAAA, a match to the Drosophila Kozak consensus sequence (C/ A)AA(A/C)ATG) of translation initiation [38], we have assigned this as the most likely start codon. Furthermore, of the preceding fourteen potential codons only two use preferred codons [39] (data not shown). The Drosophila parkin gene was reported by Greene and colleagues [40], while the current experiments were being conducted.

The parkin protein is well conserved
The Drosophila melanogaster parkin protein has been reported to be 42% identical to human parkin [40] (Fig-ure 1B). Parkin protein homologues were identified from Rattus norvegicus (NM_020093.1) and Mus musculus (AB019558.1) via the tblastn algorithm of the National Center for Biotechnology Information (NCBI). Both the R. norvegicus and M. musculus homologues were found to have 44% identity and 60% similarity to D. melanogaster parkin when analysed by the blast2 algorithm ( Figure 1B). In addition, we have determined that the Anopheles gambiae sequence XM316606.1 [41] is a homologue of parkin. Like D. melanogaster, the A. gambiae transcript has two Dipteran and mammalian parkin proteins are well conserved potential in-frame translation start sites. The Kozak sequence prior to the first ATG is very poor, however the second site closely resembles the consensus sequence and therefore it is very likely the start site. We determined that the theoretical A. gambiae parkin protein has 65% identity and 79% similarity to D. melanogaster parkin ( Figure 1B).
The parkin protein appears to be highly conserved at the amino acid sequence level.  Figure 1B). In the amino-terminal region of the proteins, the first 15 amino acids are well conserved between A. gambiae and D. melanogaster, but absent in the mammalian proteins. The human UBL shows very high similarity (62%) to human ubiquitin [17]. Correspondingly the Drosophila UBL ( Figure 1B, green box) was found to have 43% identity and 67% similarity to Drosophila ubiquitin (AAA29007; data not shown). The second highly conserved region is unique to parkin and has been termed the unique parkin domain (UPD) [42] ( Figure 1B, red box). D. melanogaster and A. gambiae share a similar eight amino acid insertion in the UPD ( Figure 1B). There are two RING-finger domains that are defined by the consensus sequence C-X 2 -C-X 9...39 -C-X 1...3 -H-X 2...3 -C/H-X 2 -C-X 4...48 -C-X 2 -C where X can be any amino acid ( Figure 1B, blue boxes) [43]. These RING-finger domains flank a cysteine-rich domain designated the In-Between Ring (IBR) domain ( Figure 1B, black box) [44]. These three domains are responsible for binding to specific E2 ubiquitin conjugating enzymes [23,28,30]. As parkin undergoes self-ubiquitination [28], conserved potential ubiquitination sites (lysine residues) were identified. There is a lysine residue that is completely conserved at K-42 of the dipterans and this corresponding residue is K-27 in mammals ( Figure 1B, black arrow). The mouse and rat parkin homologues have been recently compared to Drosophila parkin [45], however a number of the above features were not discussed. Overall parkin appears to be highly con-served between mammalian and dipteran species suggesting conservation of function among these species.
parkin suppresses degeneration of the ommaditial array in flies that express α-synuclein in the eye We generated stable transgenic flies that can conditionally express parkin when the UAS/Gal4 expression system is utilized [46]. In situ hybridization was used to confirm parkin expression in transgenic Drosophila (data not shown). Expression of parkin was directed to the developing eye using the GMR-Gal4 transgene resulting in no obvious alteration of the eye (data not shown). In vitro and cell culture research suggests that parkin can prevent α-synuclein-induced toxicity [31,47]. Expression of human α-synuclein in the Drosophila eye causes premature deterioration of the retina [32]. To examine if parkin could prevent α-synuclein-induced degeneration, we co-expressed parkin with human α-synuclein in the developing eye. Cross-sections of the retinas of one-day-old flies that express α-synuclein appear to be intact, as previously described (Figure 2A) [32]. The retinas of one-day-old flies that express both α-synuclein and parkin also appear normal ( Figure 2B,2C). As previously described, the retinas of thirty-day-old flies that express α-synuclein show signs of premature degeneration [32]. Degeneration of the normal architecture of the eye is apparent ( Figure 2D, black arrows) and reflects the disruption of the normal placement and alignment of the photoreceptors and supporting cells. In contrast, thirty-day-old flies that express α-synuclein and parkin maintain their ommatidial arrays and morphology ( Figure 2E,2F). This observation demonstrates that directed expression of parkin suppresses αsynuclein-dependent degeneration of the ommatidial array.
Retinal damage can be observed by examining an optical effect termed the pseudopupil, which is lost in aged flies that express α-synuclein [32]. We examined flies that coexpress α-synuclein and parkin, and there appeared to be some retention of this optical effect compared with flies that express α-synuclein alone (data not shown). Scanning electron microscopy of eyes revealed no obvious deterioration of the surface in flies that express α-synuclein (Figure 3A and 3D) or flies that co-express α-synuclein and parkin ( Figure 3B,3C,3E and 3F). Although α-synuclein causes degeneration of the ommatidial array, the external structure of the eye is unaffected.

Directed-expression of parkin to dopaminergic neurons does not affect climbing ability
Young wild-type adult Drosophila exhibit a strong negative geotaxis, which is increased by mechanical stimulation [48,49]. In order to measure climbing ability, flies are placed in a vial, gently tapped to the bottom and allowed to climb up the sides [32]. When parkin is expressed in the dopaminergic neurons, these flies do not show any change in their climbing ability over their life span when compared with controls ( Figure 4A). In addition, expression of parkin in dopaminergic neurons does not alter life span ( Figure 4B). These results demonstrate that parkin expression in the dopaminergic neurons has little effect upon climbing ability or life span.

Parkin suppresses α-synuclein induced loss of climbing ability
Flies that express α-synuclein, specifically in their dopaminergic neurons through the activity of the Ddc-Gal4 transgene, were assayed for their climbing ability over their life span, and were found to prematurely lose their climbing ability ( Figure 5A). Co-expression of parkin with α-synuclein suppresses this premature loss of climbing ability ( Figure 5A). This suggests that parkin can act to prevent the deleterious effects of α-synuclein expression.
Aging assays were carried out in tandem with the climbing assays described above in order to account for changes in climbing ability as a result of premature senescence.
Median survival age for flies that express α-synuclein is similar to flies that co-express α-synuclein with parkin (Figure 5B). This indicates that differences in climbing ability were not due to differences in life span.

Discussion
Drosophila parkin has a high degree of similarity to the mammalian and A. gambiae homologues. The five characteristic domains of the parkin protein, the Ubiquitin-like Domain (UBL), Unique Parkin Domain (UPD), Really Interesting New Gene finger 1 (RING1) domain, In-Between Ring (IBR) domain and RING2 all show a high degree of similarity. In addition, the two dipterans, D. melanogaster and A. gambiae, have a highly conserved extra segment of 15 amino acids at the amino-terminal of the protein. The regions between the three carboxy-terminus domains are also highly conserved, which may indicate conservation of function. Patients with ARJP caused by mutations in the UBL domain exhibit signs of lost substrate binding [23]. The UBL domain also appears to be involved in binding the Rpn10 subunit of the 26S proteasome as the R42P amino acid substitution in this domain was identified in ARJP patients and results in impaired proteasome binding of parkin ( Figure 6) [50]. Alterations of the RING1, RING2 and IBR domains of parkin result in an almost complete loss of ubiquitin conjugating enzyme H7 (UbcH7)-binding activity, which indicates that all three domains are functionally important in recruiting specific E2 ubiquitin conjugating enzymes [51]. The RING1 and RING2 domains are thought to collaborate to trap UbcH7 (Figure 6) [51]. Amino acid substitutions in the RING1 domain change the subcellular localization of parkin and enhance cytoplasmic and nuclear inclusions [52]. In addition, the amino acid substitutions C289G and C418R, which replace conserved cysteine residues in the RING domains, significantly decrease the solubility of parkin in cells [53]. Ubiquitination generally occurs near the amino-terminus of proteins and ubiquitin monomers are attached to lysine residues [25]. Several lysine residues are absolutely conserved, including one in the UBL and Expression of α-synuclein with and without parkin does not affect the external morphology of the eye two in the UPD, and these may be targets for ubiquitination. The existence of orthologues of mammalian parkin in invertebrates but not plants nor fungi [54] suggest an animal specific function for parkin activity. The highly conserved protein domains and sub-domains suggest the probable conservation of each domain's function, and given the high degree of similarity we suggest that the function of the Drosophila parkin protein is similar to that of the human parkin protein.

Expression of parkin suppresses α-synuclein-induced retinal degeneration
We demonstrate that the directed expression of parkin in the dopaminergic neurons and developing eyes leads to no obvious adverse effects. The unaltered phenotype observed when parkin is expressed in dopaminergic neurons is likely due to substrate specificity and to the ability of the parkin protein to target itself for degradation [28]. Under conditions of over-expression, parkin does not seem to target and tag essential proteins for degradation promiscuously. This may represent an excellent fail-safe mechanism the cell has developed to balance the levels of both parkin and its substrates.
The Drosophila model of ADPD has been used to examine the effect of various pharmacological agents [34,36] and other genetic aspects of the disease [35,37]. We expressed parkin along with α-synuclein and found the suppression Model of parkin directed ubiquitination of α-synuclein These results indicate that parkin may target α-synuclein for degradation in vivo ( Figure 6). Although coimmuno-precipitation studies have shown that parkin does not interact with or ubiquitinate nonmodified α-synuclein [29], parkin will ubiquitinate Oglycosylated α-synuclein [26]. Since we show suppression of the α-synuclein-induced phenotype, we believe that ectopically expressed α-synuclein is modified in Drosophila, enabling its ubiquitination by parkin. The modification of α-synuclein and subsequent ubiquitination by parkin is presented in Figure 6.
In order to select rational potential therapeutic agents, the molecular mechanisms behind disease progression must be characterized. Gene function studies with homologues of disease-causing genes in model organisms have been made practical through the advent of genome projects. Over-expression of parkin has no apparent adverse consequences and it suppresses the α-synuclein-induced PD symptoms in Drosophila. If this relationship is conserved in humans, we suggest that up-regulation of parkin should be a viable treatment for PD, and the selection of therapeutic strategies should be directed towards this end.

Conclusions
Our experiments demonstrate that the directed expression of the parkin gene counteracts the PD-like symptoms in the α-synuclein-induced Drosophila model of PD. Manipulation of the ubiquitin/proteasome degradation pathway in such a specific manner apparently remedies the toxic accumulation of α-synuclein. This study demonstrates the success of selective targeting of toxic proteins for degradation as an approach to address neurodegenerative conditions such as Parkinson's disease. The development of therapies that regulate parkin expression or parkin protein activity may be crucial in the treatment of PD.

Bioinformatic and sequence analysis
The

In situ hybridization analysis
Third instar larvae were dissected in PBS, fixed in 4% formaldehyde and dehydrated in methanol and ethanol. The carcases were probed with a DIG labelled anti-sense parkin RNA probe generated from a linear cut plasmid containing the entire parkin cDNA using the Roche Applied Science DIG Northern starter kit and reduced in size with carbonate buffer treatment. To visualize parkin RNA alkaline phosphatase labelled anti-DIG anti-bodies were incubated with the carcases and subjected to alkaline phosphate treatment as per the Roche Applied Science DIG application manual. The eye discs were dissected out completely and examined under light microscopy. The genotypes of the larvae examined were 1) w 1118 ; 2) w 1118 ; GMR-Gal4/+; UAS-parkin 1.1 /+; and 3) w 1118 ; GMR-Gal4/+; UAS-parkin 2.1 /+ and at least ten of each genotype were examined.

Aging analysis
Adult males were collected under gaseous carbon dioxide anaesthetic and aged in small groups (~10 or less per vial) upon standard cornmeal/yeast/molasses/agar media at 25°C in upright standard plastic shell vials. The flies were scored for viability every two to three days and transferred to fresh media without anaesthesia [62].

Locomotion assay
Flies were assayed for their ability to climb as described by Feany and Bender [32]. Every five days, forty male flies from a cohort of flies were assayed for their ability to climb six centimetres in eighteen seconds in a sterile plastic vial. Twenty trials were carried out for each time point. Data shown represent the results from flies tested over ninety days.