Alpha-Synuclein Disrupted Dopamine Homeostasis Leads to Dopaminergic Neuron Degeneration in Caenorhabditis elegans

Disruption of dopamine homeostasis may lead to dopaminergic neuron degeneration, a proposed explanation for the specific vulnerability of dopaminergic neurons in Parkinson's disease. While expression of human α-synuclein in C. elegans results in dopaminergic neuron degeneration, the effects of α-synuclein on dopamine homeostasis and its contribution to dopaminergic neuron degeneration in C. elegans have not been reported. Here, we examined the effects of α-synuclein overexpression on worm dopamine homeostasis. We found that α-synuclein expression results in upregulation of dopamine synthesis and content, and redistribution of dopaminergic synaptic vesicles, which significantly contribute to dopaminergic neuron degeneration. These results provide in vivo evidence supporting a critical role for dopamine homeostasis in supporting dopaminergic neuron integrity.

Expression of human a-synuclein (haSyn) in DAergic neurons of C. elegans results in their degeneration [21,22]. Yet, the effects of haSyn expression on dopamine homeostasis have not been addressed in this useful organism. Here, we used haSyn-expressing C. elegans lines to examine the toxic effects of haSyn on dopamine homeostasis and its contribution to haSyn-mediated DAergic neuron degeneration.

haSyn Expression Induces DAergic Neuron Degeneration
We first characterized the expression of dat-1 promoter-driven haSyn by using immunohistochemistry and confocal microscopy. Positive haSyn immunostaining was found exclusively in DAergic neurons, marked with dat-1 promoter-driven DsRed, demonstrating the specificity of haSyn expression in our transgenic lines ( Figure S1). Previous efforts to express wild type or pathogenic haSyn in worms led to loss of the fluorescent DAergic neuron marker due to degeneration of DAergic neurons [5,20]. Consistent with these reports, our haSyn-expressing line, but not the control line, displayed an agerelated progressive decline in the number of fluorescent DAergic neurons ( Figure 1A-E). Another haSyn-expressing line also exhibited a similar decline in the number of fluorescent DAergic neurons (data not shown). This conclusion was further confirmed by TH immunostaining experiments ( Figure S2) and similar experiments where both a nonfunctional CAT-2/TH::GFP fusion protein [23,24] and DsRed were used as DAergic neuron markers ( Figure S3).
We next investigated the effect of haSyn expression on the function of worm DAergic neurons by measuring the basal slowing response, a food-sensing behavior regulated by dopamine neurotransmission [25]. The worm basal slowing response was used to assess the effect of haSyn expression on the function of DAergic neurons [21]. As found in cat-2, a knockout mutant of worm TH, haSyn-expressing worms had an impaired basal slowing response, which returned to control levels in the presence of 0.5 mM exogenous dopamine ( Figure 1F). Thus, animals of the haSyn expressing line were functionally deficient in dopamine.
Consistent with our haSyn expression pattern, the enhanced slowing response, a food response behavior regulated by serotonin neurotransmission [25], was not affected in haSyn expressing animals ( Figure S4).
Taken together, these results lead us to conclude that haSyn expression induces degeneration of DAergic neurons in our haSyn expressing lines, similar to previous reports.

haSyn Expression Induces a Motor Capacity Deficit
We next quantified the effect of haSyn expression on worm motor capacity, which had not been assessed previously in worms specifically expressing haSyn in DAergic neurons [5,21]. In general, there are two methods to access motor capacity in worms: body bending frequency and centroid velocity [25][26][27][28]. Body bending frequency is the number of sinusoidal waves made by a worm during a given time period, while centroid velocity quantifies the physical displacement of a worm's centroid. Body bending frequency can be uncoupled from centroid displacement by genetic mutations and ageing [26,29]. We observed that L4 and day 1 adult worms exhibit similar body bending frequencies, although adult worms move much faster than L4 worms, as quantified by their centroid velocity (Cao and Feng, unpublished data). Because the centroid velocity of worm locomotion has been utilized to quantify age-related changes in motor capacity and provides more sensitive and reliable quantification of worm motor activity [26,27], this parameter was selected to address the effect of haSyn expression on the worm motor system. Indeed, haSyn expressing worms exhibited a deficit in motor activity that was restored by adding 1 mM dopamine ( Figure 2), a finding consistent with observations in a Drosophila PD model [17].

haSyn Expression Results in Altered Dopamine Metabolism
Despite their functional deficiency in dopamine neurotransmission, haSyn expressing worms surprisingly exhibited a remarkable upregulation of dopamine content from L4 to day 4 in adulthood ( Figure 3A), as measured by liquid chromatography-mass spectrometry (LC-MS). We obtained similar results and reached the same conclusion (data not shown) by using conventional high performance liquid chromatography (HPLC) as well. Consistently, the fluorescence intensity of a non-functional TH/CAT-2::GFP fusion protein [23,24] in day 2 adult haSyn expressing worms was significantly elevated ( Figure 3B).
Abnormal dopamine metabolism may produce cytotoxic molecules such as hydrogen peroxide, superoxide radicals and dopamine-quinone through two pathways, namely auto-oxidation and deamination by monoamine oxidase (MO). Dopamine deamination also yields 3,4-dihydroxyphenylacetic acid (DOPAC), a non-toxic metabolite that can be used to monitor dopamine deamination-specific oxidative stress [12,30].
We found that haSyn-expressing worms displayed an agerelated accumulation of DOPAC leading to a significantly higher DOPAC content than control worms ( Figure 3C), thereby providing evidence for an haSyn-mediated disruption of dopamine metabolism. Dopamine-quinone was not detected in any worms (data not shown), possibly because dopamine autooxidation is negligible in vivo. This quinone can be oxidized to several other species [30] or become adducted to glutathione and/ or thiol groups of native proteins [31]. Nevertheless, we conclude that haSyn expression alters dopamine metabolism in worms.

haSyn Expression Redistributes Dopamine Synaptic Vesicles
Dopamine is loaded into synaptic vesicles by a VMAT and pathogenic a-synuclein impairs dopamine storage in mammalian cell lines [32,33]. To further investigate whether haSyn expression affects dopamine homeostasis in worms, we crossed our haSyn expressing line with a worm line expressing CAT-1::GFP [34]. CAT-1 is the sole worm homolog of VMAT. In worms expressing only VMAT/CAT-1::GFP but not haSyn, the observed VMAT/ CAT-1::GFP expression pattern of DAergic neurites was continuously linear with a few bright spots at both L2 ( Figure 4A) and L4 ( Figure 4E-G) stages, a finding consistent with previous reports [34][35][36]. In contrast, many bright VMAT/CAT-1::GFP spots appeared in the remarkably weakened linear fluorescent DAergic neurites of haSyn expressing L2 worms ( Figure 4C). Such an haSyn mediated alteration of VMAT/CAT-1::GFP distribution further developed, and VMAT/CAT-1::GFP fluorescence of DAergic neurites was only located in discrete punctate spots without visible lines in L4 worms ( Figure 4I-M), which was prior to the obvious start of DAergic neuron degeneration in this worm variant.
Also consistent with previous reports [34][35][36], VMAT/CAT-1::GFP in DAergic somas of control worms was excluded from the nucleus and formed a punctate pattern in both DAergic and serotonergic neuron somas ( Figure 4B and H). haSyn expression disrupted this pattern of VMAT/CAT-1::GFP expression exclusively in DAergic but not serotonergic neurons as early as L2 ( Figure 4D, L-M). From this evidence, we conclude that haSyn expression causes dopamine synaptic vesicle maldistribution.

Disruption of haSyn-Mediated Dopamine Homeostasis Contributes to DAergic Neuron Degeneration
The next step was to determine whether haSyn-mediated disruption of dopamine homeostasis contributes to DAergic neuron degeneration in worms. In rodents, exogenous expression of DAT-1, a dopamine transporter, leads to neuronal degeneration. In worms, overexpression of TH/CAT-2 produces DAergic neuron (CEP) abnormalities [22]. Here, we found that haSyn induced DAergic neuron degeneration more slowly in worms with a cat-2 mutant background ( Figure 5), indicating that haSynmediated DAergic neuron degeneration is related to dopamine homeostasis.
Critically, VMAT/CAT-1 overexpression prohibited haSynmediated [DOPAC] upregulation ( Figure 6E), but not [dopamine] upregulation ( Figure 6D), providing evidence that enhanced sequestration of dopamine protects DAergic neurons from the toxicity of haSyn expression by affecting dopamine turnover. Thus, haSyn-mediated disruption of dopamine homeostasis significantly contributes to the observed DAergic neuron degeneration and loss of motor activity. Consistent with this conclusion, haSyn expression disturbed the VMAT/CAT-1::GFP expression pattern in L2 organisms before significant DAergic neuron degeneration starts (Figures 4 and 6B), and this disruption persisted in the cat-2 mutant background ( Figure S5), wherein DAergic neuron degeneration was prevented.

Discussion
Using in vitro and ex vivo mammalian or drosophila cell cultures, a-synuclein was found to disrupt dopamine homeostasis. Here, we provide in vivo evidence to support a critical relationship between a-synuclein and dopamine homeostasis. a-Synuclein may regulate dopamine homeostasis through multiple mechanisms [13], such as dopamine synthesis/breakdown [39,40], compartmentalization [41] and recycling [42]. Consistently, we found that a-synuclein expression altered the expression of CAT-2/TH and distribution of dopamine synaptic vesicles.
Why did we observe an haSyn mediated dopamine functional deficit along with upregulated dopamine synthesis and content? One possibility to explain this paradox is that haSyn alters dopamine synaptic vesicle trafficking or packing, which may reduce the availability of dopamine synaptic vesicles at synapses and stimulate dopamine synthesis through feedback control mechanisms [43,44]. Insufficient loading of unregulated dopamine into vesicles, therefore, could result in the observed altered dopamine metabolism. Indeed, a-synuclein was proposed to intervene directly in dopamine synaptic loading in mammals [12,32,33]. But this possibility should be further explored and validated with mammalian models.
In a previous study, investigators observed that heterological haSyn expression in worm DAergic neurons induced dopamine deficiency rather than upregulation [21]. Interestingly, haSyn expression did not cause degeneration of DAergic somas in their worm lines either. The less severe cytotoxicity of haSyn in their worm line, compared with our haSyn expressing lines and a line reported by Caldwell's group [22], may be due to different levels of protein expression.
It is worthy to point out that knockout of TH/CAT-2 or overexpression of VMAT/CAT-1 did not completely protect DAergic neurons from haSyn-mediated degeneration. Consistently, the effect of knocking out VMAT/CAT-1 on DAergic degeneration was not as pronounced as that resulting from haSyn expression, indicating that haSyn-mediated cytotoxicity is not solely caused by the disruption of dopamine homeostasis. Indeed, a-synuclein mediated modification of chaperone-mediated autophagy (CMA) also plays a critical role in DAergic neuron loss in mammals [45].

C. elegans Strains
The promoter of dat-1 was cloned and linked to a full-length cDNA encoding haSyn, DsRed or GFP according to a previous description [21]. Transgenic lines expressing haSyn were generated by injecting constructs of haSyn (10 ng/ml per injection), DsRed or GFP under the control of the dat-1 (encoding dopamine transporter) promoter sequence [46]. Two transgenic lines expressing haSyn were obtained and both exhibited similar haSyn toxicity. After the transgenic line expressing haSyn was integrated, this integrated line was backcrossed 46 with wild type worms. To produce lines expressing both haSyn and CAT-1::GFP or CAT-2::GFP, the transgenic haSyn expressing worm line was crossed into nuls26 or EM641, respectively [24,34]. All other worm protocols involved standard methods [47]. cat-1 and cat-2 mutants used were e1111 and e1112, respectively. N2 was used as the wild type.

Immunochemistry
Worms were fixed with formaldehyde and stained with goat anti-haSyn antibody according to published protocols with slight modifications [48]. All antibodies were purchased from Millipore.

Microscopy
All confocal experiments were conducted with a Leica TCS SP2 confocal microscope. The spectra used were: DsRed(l ex = 543nm and l em = 580-630nm) and GFP(l ex = 488nm and l em = 510-530nm). To count fluorescent DAergic neuron numbers, living worms were immobilized with 30 mM sodium azide on 5% agarose pads and examined with a Leica DMI3000 microscope or a Leica TCS SP2 confocal microscope according to a published method with modifications [21]. Specifically, fluorescent DAergic neurons numbers were counted manually. The existence of a fluorescent DAergic soma was evaluated by its fluorescence intensity, its position in animals and the position of its dendrites of a candidate neuron. The position of a neuron in worms and the position of its dendrites are relatively unchanged in worms throughout their life [49]. To obtain consistent data, an observer was warmed up with 10-20 day 1 animals from an integrated wild type line expressing DsRed in DAergic neurons, every day when such an experiment was conducted. These animals have eight fluorescent DAergic somas. In these experiments, representative images were captured with an Andor iXon EM 885 EMCCD camera and SimImaging (Feng, Z. unpublished software) (when Lecica DMI3000 microscope was used) or a Leica TCS SP2 confocal microscope. All images were processed and analyzed with National Instruments Vision Assistant 7.1.

Behavioral Analyses
Worm basal/enhanced slowing responses with and without dopamine pretreatment were obtained as previously described [21,25]. Locomotion speed was collected by using Automated and Quantitative Analysis of Behavior of Nematode (AQUABN) with a protocol described previously [26,27]. After a 10-minute video was collected, the average speed from minutes 7-10 was computed to eliminate the locomotion acclimation phase. For dopamine rescue experiments, dopamine was added to the liquid medium before pouring Nematode Growth Medium (NGM) plates. Animals were then raised and experiments were conducted on dopamine containing NGM plates. These dopamine-exposed, haSynexpressing animals exhibited DAergic neurite and soma degeneration phenotypes similar to haSyn expressing animals raised on regular NGM plates (data not shown).

Dopamine and DOPAC Measurements
Samples were prepared as described [21] and filtered with a 0.45 mm Millipore filter before being injected into tandem LC-MS that employed an ESI probe in the positive ion mode. The column used was a C18 Discovery HS (5mm narrow bore), 15 cm long with a 2.1 mm diameter. The mobile phase used for elution was composed of solvent A (10 mM ammonium formate, pH 3.0) and solvent B (acetonitrile) with ratios ranging from 97% -80% of solvent A. The detector was set up for single ion monitoring m/z 150-210.

Statistical Analysis
Statistical significance was analyzed by using Statistica (StatSoft, Inc.). T-tests, ANOVA with Bonferroni corrections or Dunnet's post-hoc analyses were used for their appropriate applications.