Gateways for Glutamate Neuroprotection in Parkinson’s Disease (PD): Essential Role of EAAT3 and NCX1 Revealed in an In Vitro Model of PD

Increasing evidence suggests that metabolic alterations may be etiologically linked to neurodegenerative disorders such as Parkinson’s disease (PD) and in particular empathizes the possibility of targeting mitochondrial dysfunctions to improve PD progression. Under different pathological conditions (i.e., cardiac and neuronal ischemia/reperfusion injury), we showed that supplementation of energetic substrates like glutamate exerts a protective role by preserving mitochondrial functions and enhancing ATP synthesis through a mechanism involving the Na+-dependent excitatory amino acid transporters (EAATs) and the Na+/Ca2+ exchanger (NCX). In this study, we investigated whether a similar approach aimed at promoting glutamate metabolism would be also beneficial against cell damage in an in vitro PD-like model. In retinoic acid (RA)-differentiated SH-SY5Y cells challenged with α-synuclein (α-syn) plus rotenone (Rot), glutamate significantly improved cell viability by increasing ATP levels, reducing oxidative damage and cytosolic and mitochondrial Ca2+ overload. Glutamate benefits were strikingly lost when either EAAT3 or NCX1 expression was knocked down by RNA silencing. Overall, our results open the possibility of targeting EAAT3/NCX1 functions to limit PD pathology by simultaneously favoring glutamate uptake and metabolic use in dopaminergic neurons.


Introduction
Parkinson's disease (PD) is a neurodegenerative disorder characterized by the progressive loss of dopaminergic neurons in the substantia nigra pars compacta and the development of cytoplasmic aggregation of α-synuclein (α-syn), known as Lewy bodies [1][2][3][4]. The majority of PD cases (90-95%) are believed to be sporadic and most likely caused by a complex interaction between genetic susceptibility and environmental factors. The remaining 5-10% of the cases report different version of the familial forms of PD characterized by autosomal dominant or recessive monogenic mutations in several genes such as α-syn, LRRK2, Parkin, PINK and DJ-1 [4][5][6][7][8][9][10]. Although the sequential events leading to degeneration of dopaminergic neurons is still poorly defined, many lines of evidence highlight that mitochondrial dysfunction is a common pathological feature shared by both sporadic and monogenic PD [11][12][13][14][15][16][17]. The first evidence of mitochondrial involvement in the pathogenesis of PD comes from the seminal finding that electron transfer chain complex I deficiency, induced by the mitochondrial toxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), results in an acute and irreversible parkinsonian

Cell Viability
Cell damage was evaluated by measuring the release of lactate dehydrogenase (LDH) into the culture supernatant. RA-differentiated SH-SY5Y were cultured on 12-well plates (7.5 × 10 4 cells/well) and, after being collected and centrifugated (250× g 1 min), 100 µL of supernatants was incubated with the same volume of reaction mixture (Roche Diagnostics, Monza, Italy) for 30 min at room temperature in the dark. The released LDH from dying cells was measured by reading the absorbance at 490 nm in a Victor Multilabel Counter plate reader (Perkin Elmer, Waltham, MA, USA).

Mitochondrial Activity
Mitochondrial activity was measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, which is based on the reduction of MTT into insoluble formazan by the activity of mitochondrial enzymes [54]. Briefly, RA-differentiated SH-SY5Y were cultured on 12-well plates (7.5 × 10 4 cells/well) and, at the end of the experimental protocol, cells were incubated with 1 mL of MTT reagent (0.5 mg/mL in PBS) to each well in the dark at 37 • C, in a humidified incubator, in a 5% CO 2 atmosphere. After 1 h, the supernatant was removed and centrifuged to collect all the formazan crystals, which were dissolved in 1 mL of DMSO. Finally, the absorbance was measured at 540 nm using a Victor Multilabel Counter plate reader (Perkin Elmer). Cell viability was expressed as percentage of the control.

ATP Assay
Measurement of ATP synthesis was performed by using a commercially available luciferase-luciferin system (ATPlite, Perkin Elmer) according to the manufacturer's instructions [40]. RA-differentiated SH-SY5Y cells were cultured on 96-well ViewPlate (Perkin Elmer) (1 × 10 5 cells/well) and, after being subjected to the specific treatment, intracellular ATP levels were analyzed with a luminescence counter (Victor Multilabel Counter, Perkin Elmer), normalized to the respective protein content and expressed as percentage of the control.

Detection of Mitochondrial ROS Formation
Mitochondrial ROS production was evaluated using the mitochondrial-targeted dye Mitotracker CM-H2XRos [55] (Invitrogen Life Technologies, Carlsbad, CA, USA). RA-differentiated cells were Cells 2020, 9,2037 4 of 21 cultured on glass coverslip on 6-well plates (1 × 10 5 cells/well) and, after being treated, cells were loaded with 300 nM of the dye for 30 min in the dark at 37 • C. For live cell imaging of mitochondrial ROS, measurements were conducted on the confocal microscope Zeiss 510 LSM (Carl Zeiss, Milan, Italy). CM-H2XRos was excited with 561 nm laser, and emission was detected above 580 nm and recorded for 120 s. The analysis of red fluorescence increase was performed off-line after images acquisition. Fluorescence values were reported as percentage of the control.

Analysis of Cytoplasmic and Mitochondrial Ca 2+ Levels
Cytoplasmic and mitochondrial Ca 2+ levels were monitored by single cell computer-assisted video imaging using LSM 510 confocal system (Carl Zeiss), as described previously [22]. Briefly, RA-differentiated SH-SY5Y cell were loaded with 5 µM Rhod 2-AM (Abcam, Cambridge, UK) in DMEM medium for 15 min in the dark at 37 • C; then, 4 µM Fluo-4/AM (Invitrogen Life Technologies) and 0.08% pluronic acid (Molecular Probe, Eugene, OR, USA) were added to the medium for further 45 min. Coverslips were then washed twice in PBS, placed into a perfusion chamber mounted onto the stage of an inverted Zeiss Axiovert 200 microscope, and maintained in standard buffer solution (in mM: 140 NaCl, 5 KCl, 1 CaCl 2 , 0.5 MgCl 2 , 10 HEPES, 5.5 glucose, buffered to pH 7.4 with NaOH) at 37 • C using a heated microscope stage and climate box from PeCon GmbH. Both cytoplasmic and mitochondrial Ca 2+ elevations were evaluated as fluorescence increase. Bath solution was changed with a peristaltic pump and images were acquired every 5 s. Fluo 4-AM was excited at 488 nm and the emission was time-lapsed recorded at 505-530, while Rhod 2-AM was excited at 543 nm and fluorescence emission was measured from 560 nm to 600 nm. Analysis of fluorescence was performed off-line after images acquisition, as previously described [22].

Data Analysis
Data were expressed as mean ± S.E.M. Values less than 0.05 were considered to be significant. Differences among means were assessed by one-way ANOVA followed by Dunnet's post hoc test. Statistical comparisons were carried out using the GraphPad Prism 5 software (GraphPad Software Inc., San Diego, CA, USA). Concentration-effect curves were obtained by fitting the data to the four-parameter logistic equation y = min + (max − min)/(1 + (x/EC50)n), where y is the variable of interest (normalized LDH or MTT signals), x is the α-syn concentration in nmoles/L, and n is the Hill coefficient. Data fitting was performed using the fitting routines of the Origin 8.0 software (OriginLab, Northampton, MA, USA).

Effect of α-Syn and α-Syn Plus Rot on Cell Viability
We initially exposed RA-differentiated SH-SY5Y cells to increasing concentrations of α-syn (from 3 to 30 nM) for 24 h. As shown in Figure 1A,B, we found a concentration-dependent cytotoxic effect of α-syn on both cell injury and mitochondrial activity, as assessed by LDH and MTT assays, respectively. Based on these results, we chose 10 nM as the α-syn working concentration for the following experimental sets.

Glutamate Recovery of Cell Injury, ROS Overproduction and ATP Synthesis Reduction Induced by α-Syn Plus Rot
Numerous studies showed that mitochondrial dysfunction and oxidative damage, whether triggered by environmental and/or genetic factors, contribute to the cascade of events leading to the PD neurodegeneration [21,58]. Firstly, we tested whether glutamate, as metabolic substrate, may attenuate cell damage induced by α-syn plus Rot. As expected, in RA-differentiated SH-SY5Y cells, glutamate significantly increased physiological ATP synthesis, confirming our previous results obtained in other cell models and different experimental settings [38][39][40][41] (Figure 2A). After α-syn plus Rot treatment, glutamate administration evoked an increase in ATP production up to the levels observed under control conditions ( Figure 2A). Considering that 3 µg/mL of oligomycin, an ATP synthase inhibitor [40,41], counteracted the intracellular ATP increase induced by glutamate ( Figure 2B), we suggest that this metabolic substrate may promote ATP synthesis by refilling the tricarboxylic acid (TCA) cycle and oxidative phosphorylation [39][40][41]. Furthermore, glutamate, which alone did not display any toxic effect, strongly prevented cell injury induced by α-syn plus Rot treatment ( Figure 2C,D).
Cells 2020, 9, x FOR PEER REVIEW 6 of 19 glutamate significantly increased physiological ATP synthesis, confirming our previous results obtained in other cell models and different experimental settings [38][39][40][41] (Figure 2A). After α-syn plus Rot treatment, glutamate administration evoked an increase in ATP production up to the levels observed under control conditions ( Figure 2A). Considering that 3 μg/mL of oligomycin, an ATP synthase inhibitor [40,41], counteracted the intracellular ATP increase induced by glutamate ( Figure  2B), we suggest that this metabolic substrate may promote ATP synthesis by refilling the tricarboxylic acid (TCA) cycle and oxidative phosphorylation [39][40][41]. Furthermore, glutamate, which alone did not display any toxic effect, strongly prevented cell injury induced by α-syn plus Rot treatment ( Figure 2C,D).  After investigating the ability of glutamate to stimulate the synthesis of ATP, we analyzed whether this amino acid modified ROS levels in our model. Effectively, as shown in Figure 2E,F, we found that glutamate partially prevented ROS overproduction induced by α-syn plus Rot.
which consequently affects neuronal viability [25,[61][62][63][64]. Therefore, we evaluated the effect of α-syn plus Rot on both cytoplasmic and mitochondrial Ca 2+ levels and the effect of glutamate in our experimental model. We found that, in Fluo 4-AM and Rhod 2-AM loaded RA-differentiated cells, αsyn plus Rot acute exposure led to a significant rise of both cytoplasmic ( Figure 5A,B) and mitochondrial Ca 2+ levels ( Figure 5C,D). As previously reported, under resting conditions [38,41], glutamate exposure induced a slight but significant elevation of both cytoplasmic ( Figure 5A,B) and mitochondrial Ca 2+ levels ( Figure 5C,D). Furthermore, glutamate supplementation significantly reduced both cytoplasmic ( Figure 5A,B) and mitochondrial Ca 2+ increases ( Figure 5C,D) induced by acute exposure to α-syn plus Rot.
When cells were transfected either with EAAT3 or NCX1 siRNA, glutamate failed to prevent the rise of both cytoplasmic and mitochondrial Ca 2+ levels induced by α-syn plus Rot (Figures 6 and 7).  When cells were transfected either with EAAT3 or NCX1 siRNA, glutamate failed to prevent the rise of both cytoplasmic and mitochondrial Ca 2+ levels induced by α-syn plus Rot (Figures 6 and 7).

Discussion
In the current study, we have shown that, in RA-differentiated SH-SY5Y cells exposed to α-syn plus Rot to reproduce PD-like features, glutamate supplementation significantly improved cell viability by (1) stimulating ATP synthesis, (2) reducing free radical burden, (3) limiting cytosolic and mitochondrial Ca 2+ overload. Strikingly, glutamate was completely ineffective when either EAAT3 or NCX1 expression was knocked down, disclosing an essential contribution of these transporters in mediating the effect of glutamate on mitochondrial activities. These results highlight the central role of mitochondrial dysfunctions in the pathogenesis of PD and put forward the view that progressive mitochondrial alterations can originate from bioenergetics impairments, and significantly contribute to neuronal death. On this basis, the enhancement of mitochondrial functions appears as a potential strategy to prevent, delay, or reverse PD neurodegenerative processes.
Mitochondrial activities are key determinants of cell functions, since these organelles interface with a plethora of fundamental processes, such as energy production, intracellular redox balance and cellular Ca 2+ homeostasis [65,66]. Since mitochondrial dynamics are closely integrated, any alteration taking place in one mitochondrial process can profoundly influence other functions, and therewith incite a destructive spiral of energetic failure, oxidative damage and Ca 2+ overload [12]. Several reports support the pathological relevance of this framework for the neurodegenerative processes underlying PD. For instance, postmortem studies in PD patients have consistently documented a reduction of mitochondrial complex I activity in the nigrostriatal dopaminergic neurons [20,21,[67][68][69], as well as a decrease in ATP synthesis, massive ROS production and the dysregulation of Ca 2+ homeostasis [69,70]. Similar effects can also be ascribed to α-syn and Rot exposure, which can cause a deficit of mitochondrial complex I activity as an upstream event [3,25,64,[71][72][73].
Intriguingly, it has been reported that the impairment of complex I activity could be compensated by supplying different substrates, including amino acids (such as glutamate, arginine, proline, valine, aspartate, lysine and glutamine), TCA cycle intermediates (oxaloacetate, citrate and malate) [35], and nutritional supplements like D-β-hydroxybutyrate [34], a membrane-permeable prodrug of succinate [37], or by modulating glutamate metabolism via glutamate dehydrogenase activation [36]. On this basis, we hypothesized that approaches aimed at compensating complex I deficiency may preserve the whole mitochondrial activity and positively impact cell survival in our PD model. In this regard, we focused our attention on glutamate supplementation. Under pathophysiological conditions, glutamate can be used as a metabolic substrate through its conversion to α-ketoglutarate, thereby promoting de novo ATP synthesis by entering the TCA cycle [38][39][40][41]. Here we show that, in RA-differentiated SH-SY5Y cells exposed to α-syn plus Rot, glutamate supplementation promoted cell survival. To explore the mechanisms underlying this positive effect of glutamate, we investigated some critical potentialities of this substrate. First, we verified its metabolic properties and found that, under control conditions, 1 h exposure to glutamate produced a significant increase in ATP synthesis, confirming previous findings obtained in different cell models [38] and in brain and heart isolated mitochondria [41]. When glutamate was added during the last hour of the treatment, the drop in ATP content induced by α-syn plus Rot was completely reverted. Moreover, the finding that oligomycin, an inhibitor of ATP synthase, abolished such recovery indicated that the increase in ATP synthesis induced by glutamate mainly derived from an ex novo process, which used glutamate as the main substrate.
Within inner mitochondrial membrane, complex I works in parallel with complex II, which transfers electrons from FADH 2 produced during succinate oxidation [74]. Although the contribution of this pathway to the oxidative phosphorylation is normally minimal, we hypothesized that, when complex I is compromised, glutamate can still feed the TCA cycle at the level of α-ketoglutarate, allowing dopaminergic cells to circumvent complex I and directly providing electrons from FADH 2 to complex II, finally increasing ATP generation.
Since glutamate can also participate in the synthesis of antioxidant molecules [43,75,76], we explored whether the observed protective effects could have also derived from this property. By assessing mitochondrial ROS levels, we found that, as expected, they were dramatically increased as a consequence of α-syn plus Rot treatment, and that the exposure to glutamate significantly reduced such an increase. In addition, in view of the main role of mitochondria in controlling Ca 2+ homeostasis, we also sought to explore the Ca 2+ dynamics taking place in our experimental setting. We found that the acute exposure to α-syn plus Rot induced both cytoplasmic and mitochondrial Ca 2+ alterations, which were significantly contained by glutamate addition. Overall, mitochondrial bioenergetic deficits and an increased mitochondrial oxidative damage coupled with a dysregulation of Ca 2+ homeostasis sensitized cells to intrinsic death programs, which were halted by glutamate. We did not explore the exact sequence that characterized these events. However, on the basis of the available literature, we tentatively speculate that, on the one hand, complex I alteration may trigger a bioenergetics crisis, and, on the other hand, it may favor electron leakage from the transport chain into the mitochondrial matrix, causing ROS overproduction [14,77,78]. These events initiate a vicious cycle, which may give rise to a persistent loss of both cytosolic and mitochondrial membrane potential, leading to a massive Ca 2+ influx [14,77], which in turn may fuel ROS production and ATP loss [79]. In this framework, glutamate would act by preserving the whole mitochondrial functions: by restoring energy metabolism and by improving the antioxidant defenses of the cell, it protects mitochondria from oxidative damage and ameliorates their bioenergetics performance at the same time.
Another interesting finding emerging from our study is that glutamate effect occurred in an EAAT3/NCX1-dependent fashion. We found that the protective actions of glutamate were completely lost after both pharmacological inhibition and RNA silencing of either EAAT3 or NCX1. Although in our experimental conditions RA-differentiated SH-SY5Y cells express also EAAT1-2 (data not shown) and NCX3 [40], we focused on EAAT3 and NCX1 because the molecular and functional coupling of these specific transporters has been already observed at both the plasma membrane and mitochondrial levels in different models under pathophysiological conditions [38][39][40][41]. Although the intracellular distribution of EAAT3 is less characterized than its surface membrane expression [80][81][82], being still open to further analysis, it is worth noting here that different research groups provided evidence for the localization of glutamate transporters from the EAAT family also in the mitochondrial compartment [83][84][85][86].
EAAT3/NCX1 colocalization underlies a mechanism where the entry of glutamate into the cells and mitochondria would favor the reverse mode activity of NCX, because of the increase in Na + levels that accompany the EAAT-dependent uptake of glutamate. In this way, Ca 2+ levels are influenced as well. In line with this observation, in control conditions, glutamate can induce a slight but significant increase in both cytosolic and mitochondrial Ca 2+ levels ( Figure 5), consistently with our previous findings [38,41] and as already proposed in astrocytes [87,88]. The latter event activates the mitochondrial Ca 2+ sensitive dehydrogenases, driving glutamate utilization and enhancing ATP synthesis [50]. It is worth noting that both Ca 2+ and glutamate are essential to promote the synthesis of ATP [38]. Interesting, we have previously shown that NCX3 does not support such a mechanism, since its specific silencing does not compromise glutamate-induced neuroprotection [40]. Second, a primary role of EAAT3 for the metabolic use of glutamate to fuel energetics has been already proposed in dopaminergic neurons [89]. Overall, this machinery ensures an efficient glutamate uptake, which preserves both cellular bioenergetics and antioxidant defenses. Therefore, a deficiency in their activities may compromise the intracellular utilization of glutamate, causing the loss of its protective actions. The results of the present study further support the concept that a functional interplay between these transporters is a basic requirement for proper functioning of the "glutamate machinery", otherwise known as "glutamosome" [43,84]. Furthermore, it is worth noting that, on the one hand, NCX1 knock-down may be negative since it prevented glutamate utilization; on the other hand, it may be protective against mitochondrial Ca 2+ overload in the absence of glutamate [22]. This apparent discrepancy underlies a dual role of NCX1, which operates in different ways depending on the specific molecular environment, as already observed in other pathological settings [40,90].
Experimental settings used in our work to explore and characterize the potential of glutamate-driven metabolism to limit PD-like degeneration were chosen to minimize excitotoxic neuronal damage, as discussed below. In our model glutamate, in terms of both concentration and time of exposure, had no evident toxic effects under control conditions, which is consistent with previous studies based on SH-SY5Y cells, where higher concentrations of glutamate (up to tens of millimolar levels) and/or longer exposure (within hours-to-days time frames) were reported to significantly impair cell functions/viability [91][92][93][94][95]. Several lines of evidence converge to support that the metabolic fate of glutamate in brain cells is highly influenced by the glutamate levels in the extracellular milieu: glutamate metabolism that feeds ATP-generating pathways is typically favored when neuronal (including mesencephalic cell preparations) or glial cultures are exposed for 1-2 h to glutamate concentrations in the hundreds of micromolar range [40,89,[96][97][98]. Similar levels can be transiently reached following glutamate release within extracellular compartments at the tripartite synapse structures [99,100] and do not significantly affect viability when primary neuronal, glial or cardiac cell cultures are similarly exposed to glutamate for 1-3 h [39,89,96,101,102]. Although in our model EAAT/NCX transporters emerge as a protective gateway that contributes to cellular uptake and metabolism of glutamate, three major complexities remain to be further addressed. First, with the intent to move from in vitro to in vivo studies, the characterization of the EAAT/NCX pathway for glutamate use in PD cannot simply rely on exogenous glutamate administration, since this amino acid does not easily cross an intact blood-brain barrier [99,103] and/or can provoke undesirable excitotoxic insults [42,104]. To this end, an approach aimed at modulating EAAT/NCX functional interaction would be more feasible to safely influence both extracellular glutamate levels and its intracellular metabolic use. It is worth mentioning here that other interventions that similarly favor glutamate metabolism were shown to improve cell bioenergetics as well as promote clearance of extracellular glutamate and survival of degenerating neurons after an ischemic insult [105,106]. Second, we cannot rule out a possible involvement of metabotropic receptors (mGluRs) for the glutamate benefits observed in our models, since mGluRs have been detected in RA-differentiated SH-SY5Y cells [91,107]. Third, the relevance of EAAT/NCX functional interaction in PD pathology needs to be also explored in glial cells, which normally contribute to the maintenance of brain homeostasis [108], can metabolically respond to glutamate in an EAAT/NCX-dependent manner [38,41] and ultimately can play a relevant role in degenerative disorders, including PD [109][110][111].
Collectively, results from the present study add further ground to the hypothesis that mitochondrial dysfunctions are major drivers of the neurodegenerative cascade underlying dopaminergic neuronal death in PD. The modulation of EAAT/NCX function may be an effective step forward in tackling the challenge of PD neurodegeneration, since it has the potential of simultaneously address both glutamate-mediated excitotoxicity (by favoring cell glutamate uptake) and the upstream energetic failure (by favoring glutamate metabolic use) that incite the progression of PD pathology. In this regard, further studies are required to understand the specific role of EAAT3-NCX1 functions in contributing to glutamate metabolic use in animal models of PD.