E46K mutant α-synuclein is more degradation resistant and exhibits greater toxic effects than wild-type α-synuclein in Drosophila models of Parkinson's disease

Parkinson’s disease (PD) is one of the most common neurodegenerative diseases, which is characterized by progressive motor dysfunction as well as non-motor symptoms. Pathological and genetic studies have demonstrated that α-synuclein (αSyn) plays key roles in the pathogenesis of PD. Although several missense mutations in the αSyn gene have been identified as causes of familial PD, the mechanisms underlying the variance in the clinical phenotypes of familial PD caused by different mutations remain elusive. Here, we established novel Drosophila models expressing either wild-type (WT) αSyn or one of five αSyn mutants (A30P, E46K, H50Q, G51D, and A53T) using site-specific transgenesis, which express transgenes at equivalent levels. Expression of either WT or mutant αSyn in the compound eyes by the GMR-GAL4 driver caused mild rough eye phenotypes with no obvious difference among the mutants. Upon pan-neuronal expression by the nSyb-GAL4 driver, these αSyn-expressing flies showed a progressive decline in locomotor function. Notably, we found that E46K, H50Q, G51D, and A53T αSyn-expressing flies showed earlier onset of locomotor dysfunction than WT αSyn-expressing flies, suggesting their enhanced toxic effects. Whereas mRNA levels of WT and mutant αSyn were almost equivalent, we found that protein expression levels of E46K αSyn were higher than those of WT αSyn. In vivo chase experiments using the drug-inducible GMR-GeneSwitch driver demonstrated that degradation of E46K αSyn protein was significantly slower than WT αSyn protein, indicating that the E46K αSyn mutant gains resistance to degradation in vivo. We therefore conclude that our novel site-specific transgenic fly models expressing either WT or mutant αSyn are useful to explore the mechanisms by which different αSyn mutants gain toxic functions in vivo.


Introduction
Parkinson's disease (PD) is one of the most common neurodegenerative diseases, and is characterized by progressive motor dysfunction, such as resting tremor, bradykinesia, and rigidity, as well as non-motor symptoms, including olfactory deficit, autonomic dysfunction, and sleep disturbance. The pathological hallmark of PD is the loss of dopaminergic neurons in the substantia nigra, accompanied by the deposition of intraneuronal inclusions called Lewy bodies (LBs), which is comprised mainly of α-synuclein (αSyn). Although the majority of PD cases are sporadic, about 10% of cases are familial, and both missense and multiplication mutations in the αSyn gene (SNCA) were discovered to cause familial PD [1][2][3][4]. Moreover, genome-wide association studies identified single-nucleotide polymorphisms (SNPs) in the SNCA gene to be major risk factors for sporadic PD [5,6]. Considering these pathological and genetic findings, αSyn is thought to play key roles in the pathogenesis of PD.
Several missense mutations of αSyn that are responsible for familial PD have been identified so far, including A30P, E46K, H50Q, G51D, A53E, and A53T [1,[7][8][9][10][11]. However, how these different mutations contribute to the pathogenesis of PD still remains elusive. In previous in vitro studies, E46K, H50Q, and A53T αSyn have been shown to have higher aggregation propensity than wild-type (WT) αSyn, whereas A30P and G51D αSyn have lower aggregation propensity [12][13][14][15][16]. On the contrary, in vivo studies focusing on the aggregation-resistant tetramer and aggregation-prone monomer forms of αSyn reported that A30P, E46K, H50Q, G51D, and A53T mutations decreased tetramer:monomer ratios in cell culture and mouse brains [17][18][19]. E46K αSyn, but not A30P or A53T αSyn, was also reported to show enhanced phosphorylation of the Ser-129 residue in human cells, yeast, and mouse brains [20]. Considering the prominent importance of αSyn in the pathogenesis of PD, elucidating the pathomechanisms by which αSyn mutations gain neurotoxicity is indispensable to understand PD pathogenesis. To elucidate the pathological effects of αSyn mutations, we established transgenic Drosophila models of PD expressing WT αSyn or αSyn mutants using site-specific transgenesis, by which the transgene is inserted into the same locus of the genome, and thus the transgenes are expected to be expressed at equivalent levels [21,22]. This method enables us to precisely compare the effects of each mutation in vivo. Using these transgenic Drosophila lines, we showed that the neuronal expression of E46K, H50Q, G51D, and A53T αSyn in flies results in stronger toxic effects than the expression of WT αSyn. We found that the protein expression level of E46K αSyn was higher than that of WT αSyn, despite equivalent mRNA expression levels. Furthermore, we demonstrated through in vivo chase experiments that degradation of the E46K αSyn protein was significantly delayed compared with WT αSyn. These results imply that one of the pathological effects of the E46K mutation in PD pathogenesis is conferring resistance to degradation.
The sequence of primers for site directed mutagenesis are as follows:

Fly eye imaging
Light microscope observation of fly eyes was performed using a stereoscopic microscope (SZX10, Olympus, Tokyo, Japan) with a digital camera unit (DP21, Olympus). Scanning electron microscopic (SEM) images were taken using an electron microscope (TM1000, Hitachi, Tokyo, Japan). One-day-old male adult flies were used.

Climbing assay
The climbing assay was performed according to a published protocol [26]. Six to twenty male flies were used for each genotype. Climbing scores were obtained from four to five independent experiments.

Immunoblotting
Fly heads were homogenized in Triton lysis buffer (50 mM Tris-HCl, pH 7.4, 1% Triton X-100, 150 mM NaCl, and 1 mM ethylenediaminetetraacetic acid (EDTA)) containing a protease inhibitor mixture (cOmplete, EDTA-free, Roche Applied Science, Indianapolis, IN) and centrifuged at 15,000 g for 20 min at 4˚C, and the supernatants were collected as the Tritonsoluble fractions. For the preparation of Triton-insoluble fractions, the remaining pellets were washed twice with Triton lysis buffer and lysed in sodium dodecyl sulfate (SDS) buffer (2% SDS, 90 mM Tris-HCl pH 6.8, 20% glycerol). For unfractionated samples, fly heads were homogenized in SDS buffer, centrifuged at 12,000 g for 5 min at 4˚C, and the supernatants were collected. The proteins were separated by 5%-20% polyacrylamide gels, transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad Laboratories, Hercules, CA) and incubated with phosphate-buffered saline containing 0.4% paraformaldehyde for 30 min at room temperature [27] before blocking with PVDF Blocking Reagent for Can Get Signal (TOYOBO Co., Ltd., Osaka, Japan). The antibodies used in this study were as follows: anti-αSyn (clone 42, BD Transduction Laboratories, Franklin Lake, NJ), anti-actin (clone AC40, Sigma-Aldrich, St. Louis, MO) at 1:1,000 dilution, and HRP-conjugated secondary antibodies at 1:10,000 dilution (Jackson ImmunoResearch Laboratory, West Grove, PA). The bands were visualized with ImmunoStar Zeta (Wako Pure Chemical Industries, Osaka, Japan), and images were captured by an Asherman Imager 600 (GE Healthcare Life Science, Pittsburgh, PA). Signal intensities were quantified by densitometry using ImageJ v1.50i software (National Institutes of Health, Bethesda, Maryland).

Quantification of αSyn protein turnover
To assess αSyn protein turnover, the GeneSwitch conditional expression system was used, as previously described [28]. Briefly, RU486 (mifepristone, Sigma-Aldrich) was dissolved in 100% ethanol, further diluted in water, and then mixed with Instant Drosophila medium (Carolina Biological Supply Company, Burlington, NC) at a final concentration of 10 μg/mL. For RU486 treatment, flies were raised on RU486-containing medium from the larval stage until adulthood. After eclosion, the flies were put on standard cornmeal medium for the indicated time periods and fly heads were homogenized in SDS buffer and centrifuged at 12,000 g for 5 min at 4˚C. The supernatants were then subjected to immunoblot analysis, as described above.

Statistical analyses
Data were analyzed using Excel 2007 (Microsoft, Redmond, WA) or R version 3.5.2 (The R Foundation for Statistical Computing, Vienna, Austria). One-way ANOVA followed by the Dunnett's post hoc test was used to analyze differences in climbing scores, αSyn mRNA expression levels, and αSyn protein expression levels. The two-tailed Student t-test was used to identify differences in the degradation chase experiments.

Drosophila models of PD expressing αSyn and its mutants at comparable levels
To analyze how familial PD mutations affect αSyn toxicity in vivo, we generated transgenic Drosophila lines expressing WT or mutant forms (A30P, E46K, H50Q, G51D, and A53T) of human αSyn (Fig 1A). Instead of conventional random transgenesis, we employed the sitespecific transgenesis system using phiC31 integrase, which mediates sequence-directed integration of transgenes in a distinct genomic locus [21,22]. This method enables us to precisely compare the effects of each mutation, because the expression levels of αSyn are expected to be consistent and position effects of transgenes could be avoided. Indeed, quantitative RT-PCR analysis showed that neither of the mutant αSyn flies showed differences in αSyn mRNA levels compared with WT αSyn flies under the control of the eye-specific GMR-GAL4 driver or the pan-neuronal nSyb-GAL4 driver (Figs 1B and 1C). We also confirmed by quantitative RT-PCR that αSyn mRNA levels of all six transgenic flies without the GAL4 driver were less than 1% of the levels in flies with the nSyb-GAL4 driver (S1 Fig). This result demonstrates that these newly established transgenic flies produce mRNA transcripts of αSyn at comparable levels. We noted that the αSyn mRNA levels of our flies are 1.5-fold higher than those of the flies previously established by random transgenesis (denoted as WT(R) in S2 Fig) [25].

Eye-specific expression of WT and mutant forms of αSyn causes comparable compound eye degeneration
It has been reported that the overexpression of WT αSyn in the compound eyes of flies causes the rough eye phenotype as well as abnormal morphology of ommatidia, both of which are readily analyzed by microscopic observation [29][30][31][32]. To compare the toxicity of αSyn with various familial mutations, we analyzed the compound eye morphology of flies expressing either WT or mutant αSyn. Light microscopic observation demonstrated that WT αSyn flies showed neither an apparent rough eye phenotype nor loss of pigmentation, similarly to the  Fig 2B). These results demonstrate that eye-specific expression of WT αSyn causes weak degeneration in the compound eyes. The flies expressing αSyn with familial PD mutations also showed abnormal ommatidial morphology and bristle patterns, but not an apparent rough eye phenotype nor loss of pigmentation (Fig 2A and 2B). These flies showed eye degeneration at almost similar levels, and no apparent differences in the severity of eye degeneration were observed when compared with the WT αSyn flies. These results indicate that αSyn expression in the compound eyes causes weak degeneration, the extent of which would not be significantly affected by familial mutations.

Neuronal expression of A53T, E46K, G51D, and H50Q αSyn causes an earlier decline in locomotor function than WT αSyn
We next analyzed whether the familial mutations affect the toxicity of αSyn against neuronal functions. It has been reported that neuron-specific expression of WT αSyn in flies causes neuronal dysfunction, which can be detected as the progressive decline of locomotor function [33]. Therefore, we generated flies expressing different forms of αSyn under the control of a pan-neuronal nSyb-GAL4 driver and analyzed locomotor function by the climbing assay. Quantitative analysis of the climbing scores demonstrated that the locomotor function of WT αSyn flies progressively decreased by 7 weeks of age, which is a much faster decline than that of the control flies bearing nSyb-GAL4 alone (Fig 3A). This result indicates that nSyb-GAL4mediated expression of WT αSyn causes progressive locomotor dysfunction, which is in good agreement with previous reports in which αSyn was expressed by different neuron-specific drivers, such as elav-GAL4 [33][34][35][36].
Flies expressing mutant αSyn in neurons also showed a progressive decline in locomotor function (Fig 3A). Interestingly, flies expressing either A53T, E46K, G51D, or H50Q αSyn  showed a significant decrease in climbing scores compared with control flies from 3 weeks of age, whereas flies expressing WT αSyn did not (Fig 3B). At 5 weeks of age, all mutant αSynexpressing flies except for A30P αSyn-expressing flies showed decreased locomotor function compared with control flies (Fig 3C), although a significant difference was not detected compared with WT αSyn-expressing flies. These results suggest that WT αSyn and its familial mutants cause locomotor dysfunction upon their neuronal expression, and that several forms of mutations, such as A53T, E46K, G51D, and H50Q enhance αSyn toxicity in vivo.

E46K αSyn mutant accumulates in vivo
Although flies expressing A53T, E46K, G51D, and H50Q αSyn showed an earlier decline in locomotor function, how these mutations accelerate αSyn toxicity remains to be elucidated. Because increased levels of αSyn have been reported to accelerate the onset and progression of disease-associated phenotypes in patients [2][3][4], we analyzed the protein levels of αSyn in our flies. Immunoblotting analysis of Triton X-100-soluble fractions of eye-specific αSyn-expressing fly homogenates revealed that E46K αSyn-expressing flies showed a 63% increase in αSyn protein level compared with WT αSyn-expressing flies, whereas the other mutant flies showed no differences (Fig 4A and 4B). We noted that αSyn protein levels of our flies are 1.   analysis of total fractions of pan-neuronally αSyn-expressing fly homogenates also showed a 53% higher αSyn level in E46K αSyn flies than in WTαSyn flies (Figs 4C and 4D). These results suggest that the E46K mutation results in increased monomer protein levels of αSyn, which might be a reason for the accelerated toxicity of this mutation in vivo.

E46K mutation of αSyn gains resistance to degradation in vivo
We then analyzed the molecular mechanism underlying how the E46K mutation leads to αSyn accumulation in flies. In general, increased cellular protein levels are caused by increased rates of transcription and translation from each gene, or by delayed rates of protein degradation. Because mRNA levels of the αSyn transgenes are almost consistent among the flies generated in this study (Fig 1B), levels of the αSyn proteins are assumed to be comparable. Therefore, we hypothesized that the E46K mutation might delay αSyn protein degradation, leading to its accumulation in vivo.
To test this hypothesis, we employed an inducible expression system using GeneSwitch, in which expression of the genes of interest can be regulated by the addition of mifepristone (RU486) [37]. Using this system, we induced the expression of either WT or E46K αSyn only at the larval stage, and examined the rate of decrease in αSyn protein level after eclosion by immunoblotting analysis of adult flies collected at different time points (Fig 5A). Inducible αSyn expression was confirmed by comparing the amount of αSyn in flies raised in medium containing RU486 with those raised without RU486, although low levels of leak expression were observed (Fig 5B). Interestingly, whereas αSyn protein levels in WT αSyn-expressing flies decreased linearly from eclosion to 7 days of age, levels in E46K αSyn-expressing flies remained at the initial level even at 4 days of age, followed by a decline at 7 days of age ( Fig  5C). This result indicates that E46K αSyn shows delayed degradation compared with WT αSyn, which would lead to increased levels of E46K αSyn in vivo.

Discussion
In this study, we established novel Drosophila models expressing WT αSyn or αSyn mutants using site-specific transgenesis, which express transgenes at equivalent levels. We showed that flies expressing either E46K, H50Q, G51D, or A53T αSyn show earlier onset of locomotor dysfunction than flies expressing WT αSyn. We found that the expression level of the E46K αSyn protein was higher than that of WT αSyn, despite equivalent mRNA expression levels. In vivo chase experiments demonstrated that degradation of the E46K αSyn protein was significantly delayed compared with WT αSyn, indicating that the E46K αSyn has higher resistance to degradation than WT αSyn in vivo.
Drosophila models of PD expressing not only WT αSyn but also some familial PD-linked forms of mutant αSyn, such as A30P, E46K, H50Q, G51D, and A53T, have been previously established by random transgenesis [25,33,38]. A30P αSyn-expressing flies were reported to demonstrate stronger phenotypes than WT αSyn-expressing flies, although expression levels of the transgenes were not analyzed [33]. Mohite et al. also established flies expressing each of E46K, H50Q, and G51D by random transgenesis, and analyzed the fly lines with equivalent protein expression levels [38]. Their E46K, H50Q, and G51D αSyn-expressing flies showed more severe declines in locomotor function than WT αSyn-expressing flies. This is in agreement with our results, implying that these αSyn mutants have increased toxic effects compared with WT αSyn. However, considering that these fly lines were established by random transgenesis [39], they are expected to have different transgene integration sites in the genome, and thus the possibility of position effects cannot be excluded. Therefore, we here used phiC31 integrase-mediated site-specific transgenesis [21] to establish transgenic Drosophila lines E46K αSyn mutation confers degradation resistance in vivo carrying a single copy of each WT or mutant αSyn transgene in the same genomic locus, which has advantages for directly comparing the effects of different αSyn mutants. In this study, we successfully found a delayed decay of E46K αSyn using our new αSyn transgenic flies established by site-directed transgenesis. Thus, our PD fly models are the first in vivo models that are suitable for studying the pathological effects of αSyn mutants.
We examined the effects of αSyn expression in the compound eyes and nervous system of flies. Although αSyn expression in the compound eyes showed mild phenotypes, phenotypic differences among the mutants were not apparent, probably because of the limited sensitivity of quantitative phenotypic evaluation. To quantify the phenotypic differences among the mutants more sensitively, we analyzed their locomotor functions upon neuronal expression of αSyn. E46K, H50Q, G51D, and A53T αSyn-expressing flies demonstrated earlier declines in their climbing scores than WT αSyn-expressing flies. It is noteworthy that familial PD patients carrying E46K, H50Q, G51D, or A53T mutations were reported to develop more severe clinical phenotypes than typical PD, including dementia or cognitive dysfunction [1,8,10], although the number of clinical reports of familial PD patients is limited. A30P αSyn-expressing flies did not show an earlier onset of the climbing phenotype than WT αSyn-expressing flies, which is inconsistent with a previous report [33]. However, since the wild-type and A30P-expressing flies in the previous report were generated by random transgenesis, this discrepancy may be simply owing to differences in protein expression levels and/or position effects of the transgenes. Therefore, severe phenotypes in our E46K, H50Q, G51D, and A53T αSyn-expressing flies may reflect the severe clinical phenotypes of familial PD patients.
Several studies have explored the pathomechanisms of these mutant forms of αSyn responsible for familial PD. Since αSyn aggregates in vitro, and accumulates as Lewy bodies in patients' brains, the aggregation propensity of αSyn mutants has been studied extensively. In vitro studies showed higher aggregation propensities of E46K, H50Q, and A53T αSyn than WT αSyn, as shown by thioflavin-T assays [13][14][15][16]. On the contrary, G51D αSyn has been reported to have a lower aggregation propensity than WT αSyn [12]. It is also reported that αSyn interacts with membrane lipid components, which could be targets for αSyn toxicity [40][41][42]. In vitro studies showed that A30P and G51D mutations of αSyn had decreased lipid binding, whereas A53T and H50Q mutations did not differ from WT αSyn in their lipid binding [43][44][45][46]. Therefore, a universal mechanism of enhanced toxicity that applies to all αSyn mutants still remains elusive, and each mutation may have different and multiple mechanisms for their toxic effects.
In this study, we found that expression levels of the E46K αSyn protein were significantly higher than those of the WT αSyn protein in our fly models, despite their equivalent mRNA levels. We further showed using in vivo chase experiments that degradation of the E46K αSyn protein was slower than that of the WT αSyn protein. A previous study using optical pulsechase experiments reported that the half-life of the Dendra2-tagged E46K αSyn protein did not differ from that of the Dendra2-tagged WT αSyn protein in rat cortical neurons, although possible effects of the Dendra2-tag on αSyn turnover could not completely be excluded [47]. More recently, the E46K αSyn protein was reported to be degraded by both the proteasome and the macroautophagy pathway in PC12 cells, and cycloheximide chase experiments showed that degradation of the E46K αSyn protein was slower than that of the WT αSyn protein, consistent with our results [48], although we did not exclude the possible dysfunction of background degradation systems by expression of E46K αSyn. Considering that the WT αSyn protein was reported to be degraded by both the proteasome and chaperone-mediated autophagy pathway [49] [50], and that A30P and A53T αSyn mutants were reported to be resistant to degradation by chaperone-mediated autophagy [50], gaining resistance to degradation may play important roles in the pathogenesis of familial PD. Taking advantage of their suitability for genetic analyses, our PD fly models expressing WT and mutant αSyn at equivalent levels, which were generated by site-specific transgenesis, are useful in vivo models to study the pathomechanisms of PD.