(Poly)phenols protect from α-synuclein toxicity by reducing oxidative stress and promoting autophagy

Abstract

Parkinson's disease (PD) is the most common movement neurodegenerative disorder and is associated with the aggregation of α-synuclein (αSyn) and oxidative stress, hallmarks of the disease. Although the precise molecular events underlying αSyn aggregation are still unclear, oxidative stress is known to contribute to this process. Therefore, agents that either prevent oxidative stress or inhibit αSyn toxicity are expected to constitute potential drug leads for PD. Both pre-clinical and clinical studies provided evidence that (poly)phenols, pure or in extracts, might protect against neurodegenerative disorders associated with oxidative stress in the brain. In this study, we analyzed, for the first time, a (poly)phenol-enriched fraction (PEF) from leaves of Corema album, and used in vitro and cellular models to evaluate its effects on αSyn toxicity and aggregation. Interestingly, the PEF promoted the formation of non-toxic αSyn species in vitro, and inhibited its toxicity and aggregation in cells, by promoting the autophagic flux and reducing oxidative stress. Thus, C. album (poly)phenols appear as promising cytoprotective compounds, modulating central events in the pathogenesis of PD, such as αSyn aggregation and the impairment of autophagy. Ultimately, the understanding of the molecular effects of (poly)phenols will open novel opportunities for the exploitation of their beneficial effects and for drug development.

Introduction

Parkinson's disease (PD) is the second most common age-related neurodegenerative disease after Alzheimer's disease (1). Despite intensive studies on the molecular basis of this disorder, we still lack a comprehensive understanding of the underlying mechanisms, compromising the development of effective therapeutic strategies. PD is deeply associated with the misfolding and aggregation of α-synuclein (αSyn), a presynaptic protein whose aggregation has also been linked to other disorders known as synucleinopathies (2,3). Thus, proteostasis dysfunction is central in PD and includes proteasomal, lysosomal and autophagic dysfunction. Although different pathways have been shown to contribute to the degradation of αSyn (4), autophagy is important in the clearance of αSyn aggregates (5,6), and the proteasome is involved in the degradation of soluble and small oligomeric αSyn species (7,8).

Currently, oligomeric αSyn species, rather than insoluble aggregated forms, are thought to be associated with αSyn cytotoxicity (9,10). The process of αSyn aggregation can be modulated by various internal and external stimuli, including oxidants and mitochondrial toxins (11), αSyn is highly susceptible to oxidative stress-induced posttranslational modifications that induce its misfolding and cytotoxicity (12,13). Moreover, oxidized αSyn is more prone to aggregation (14) and, on the other hand, aggregated αSyn induces oxidative stress (15), supporting a prominent role for oxidative stress in the onset and progression of PD. In addition, dopamine metabolism in the brain is also known to contribute to high levels of oxidative stress and neurodegeneration in PD (2,3). Therefore, the identification of novel agents that either prevent the production of reactive oxygen species (ROS) or inhibit αSyn aggregation are thought to be of particular interest.

(Poly)phenols, either as dietary (poly)phenols or nutraceutical supplements, are known for their protection against oxidative damage in the brain (16–18), suggesting that they may be useful modifiers of αSyn aggregation (19). The term phenol is used to describe a structure with at least one aromatic ring with one or more hydroxyl groups attached (20). Flavonoids, with several aromatic rings, are known as polyphenols. However, other compounds, including hydroxycinnamates and phenolic acids, with only one phenolic ring are also referred to as polyphenols. Therefore, polyphenols will be hereafter referred to as (poly)phenols or (poly)phenolic compounds. These phytochemicals occur naturally in plants and epidemiological, pre-clinical and clinical studies showed their importance for human health as they reduce the incidence and prevalence of cardiovascular diseases, cancer, diabetes, inflammation and age-related disorders (16,19,21).

(Poly)phenols were initially thought to protect cells against oxidative damage through the scavenging of free radicals. However, several other mechanisms underlying the neuroprotective action of (poly)phenols have been recently described, such as iron chelation, regulation of calcium homeostasis, activation of signaling cascades, promotion of amyloid precursor protein processing and reduction of β-amyloid fibril formation (19,22).

Several cellular models have been extensively used to study the molecular basis of different neurodegenerative disorders (23–25). In yeast, expression of human αSyn induces toxicity and inclusion formation in a concentration-dependent manner. These inclusions consist of both oligomeric and higher-order aggregated species of αSyn (25,26). Impairment of cellular quality control systems, the generation of ROS and apoptosis, along with other features typically associated with neurodegeneration were also identified in yeast (27).

Here, we tested (poly)phenol-enriched fractions (PEFs) from leaves or fruits of Arbutus unedo and Corema album against αSyn aggregation and toxicity in yeast and human cell models of PD (24,28). Leaves of Ginkgo biloba and fruits of Rubus idaeus were also analyzed due to their known beneficial effects (29–32). Arbutus unedo is a native Mediterranean species with reported antioxidant and beneficial health properties (31,33,34), while the bioactivities from C. album native from Iberia Peninsula are poorly characterized, representing a pool of phytochemicals yet to be explored.

This study demonstrates that (poly)phenols from leaves and fruits of A. unedo and C. album protects yeast cells from general oxidative stress and also from αSyn-induced toxicity, two major hallmarks of PD. Interestingly, PEF from C. album leaves, characterized for the first time in this study, were the most promising of the PEFs tested. This presented a higher protective capacity than G. biloba, a species recognized for its beneficial health properties (32). Here, we show that the mechanisms underlying C. album leaf PEF protection against αSyn-induced toxicity were associated with its ability to stabilize αSyn non-toxic oligomeric species in vitro, reduce ROS levels and increase the autophagic function, which is compromised by αSyn expression. These findings suggest that C. album leaf (poly)phenols are putative therapeutic agents for preventing αSyn toxicity in PD and other synucleinopathies.

Results

Chemical characterization of PEFs

PEFs, clean of organic acids, sugars and minerals, were obtained from fruits and leaves of A. unedo and C. album, leaves of G. biloba, and fruits of R. idaeus (the latest two used as a reference due to their well-established properties) (29–32). PEFs were then assessed for total phenolic content and in vitro antioxidant capacity for peroxyl radical (Table 1), a ROS with biological relevance in human physiology (35). Corema album leaf PEF presented the higher antioxidant activity; however, A. unedo fruit PEF presented the higher antioxidant potential (antioxidant capacity per total phenol content) (Table 1).

Table 1.

Total phenol content of PEFs and in vitro peroxyl radical scavenging activity

Matrix	Species	Antioxidant capacitya (µmol TE g−1 dw)	Total phenol contenta (mg GAE g−1 dw)	Antioxidant capacity per total phenol contenta (µmol TE mg−1 GAE)
Leaves	Ginkgo biloba	2162.94 ± 207.32	88.83 ± 14.00	24.35 ± 1.50
	Arbutus unedo	2331.37 ± 212.35 n.d.	121.77 ± 11.25**	19.15 ± 0.02***
	Corema album	2886.19 ± 151.65***	112.35 ± 3.96*	25.69 ± 0.44 nd
Fruits	Rubus idaeus	444.57 ± 19.25	13.64 ± 0.09	32.60 ± 1.20
	Arbutus unedo	1498.36 ± 52.46***	10.49 ± 0.08***	142.78 ± 3.90***
	Corema album	956.19 ± 54.46**	15.81 ± 0.12***	60.48 ± 3.09***

^aValues represent the mean ± SD of three independent experiments. The values of each matrix were statistically compared with the value of the respective comparison species (G. biloba for leaves and R. idaeus for fruits). Statistically significant differences are shown, *P < 0.05, **P < 0.01, ***P < 0.001 and n.d. as not different for P < 0.05.

The differences observed in the PEFs antioxidant capacity arise from its chemical composition. In order to identify the main compounds present in the plants under study, PEFs were analyzed by HPLC-PDA-ESI-MS (Supplementary Material, Fig. S1 and Tables S1–S6).

Regarding the PEFs from leaves, the main components of G. biloba were flavonol glycosides, biflavonoids, terpenes and flavones, in agreement with previous studies (36). Arbutus unedo leaves contained mainly gallic acid derivatives (37), catechin, tannins and flavonol derivatives. Corema album leaves presented flavonol derivatives, especially myricetin, (epi)catechin and proanthocyanidins. It also contained a range of late-eluting components with MS properties suggesting to be stilbene derivatives (Supplementary Material, Fig. S1).

Arbutus unedo fruits yielded mainly gallic acid derivatives (also common in this species leaves) (37), conjugated with quinic acid, shikimic acid or glucose. Moreover, anthocyanins, quercetin and ellagic acid derivatives were identified, as in (31,37,38). Rubus idaeus fruits were dominated by anthocyanins and ellagitannins (Sanguiin H6 and Lambertianin C), with a number of other minor components (38,39). Corema album fruits, still poorly characterized, were mainly composed of derivatives of chlorogenic acid and flavonols, in agreement with the literature (37,40). The main differences obtained between this fruit and the other two analyzed was the absence of anthocyanins, ellagic acid or ellagitannins and the presence of myricetin, kaempferol and caffeic acid derivatives.

(Poly)phenols reduce the formation of αSyn inclusions and its cytotoxicity in yeast

We used an established yeast model of αSyn aggregation and toxicity (28) to screen for PEFs ability to modulate αSyn toxicity. We monitored cell growth (Fig. 1A), determined cell metabolic capacity (Fig. 1B) and quantified the percentage of cells displaying αSyn inclusions (Fig. 1C). Yeast cells pre-treated with PEFs for 6 h were further incubated with PEFs in glucose (control) or galactose medium, where αSyn expression is induced.

Figure 1.

PEFs protect yeast from αSyn-induced toxicity and reduce the percentage of cells with αSyn inclusions. BY4741 cells transformed with the αSyn encoding plasmid were treated with the indicated PEFs for 6 h in raffinose medium, prior to treatment with PEFs in glucose (control, αSyn OFF) or galactose medium (αSyn ON). (A) Growth curves of cells subjected to the described treatment for 24 h: control (filled square), αSyn (open square) or αSyn with PEFs (gray filled circle).Values represent the mean ± SD of at least three independent experiments. (B) Cells metabolic capacity, expressed in arbitrary units (×1000), assessed at 6 h after αSyn expression induction. (C) Percentage of cells with αSyn inclusions, determined by fluorescence microscopy imaging, after 6 h of αSyn-GFP expression induction. (D) Viability of cells subjected to the treatment with C. album leaf PEF (filled square) or the pure compounds myricetin-3-O-galactoside (open square), myricetin-3-O-glucoside (filled circle) or quercetin-3-O-galactoside (gray filled triangle), at the indicated concentrations. Viability was evaluated by percentage of PI-positive cells normalized to αSyn untreated cells (fold increase), assessed by FCM. Values represent the mean ± SD of at least three independent experiments. Statistically significant different results from cells expressing αSyn not submitted to PEFs treatment are shown, *P < 0.05, **P < 0.01 and ***P < 0.001.

The concentrations tested were selected based on the analysis of the growth curves of BY4741 (WT, control) strain and the determination of the lag and doubling time of cells treated with various concentrations of PEFs (Supplementary Material, Fig. S2). Δyap1, Δsod1 and AD1–8 strains were also studied for comparison, as these strains bear deletions in different genes related to oxidative stress tolerance and compound permeability (31,38) (Supplementary Material, Fig. S2). At the lower concentration, the PEFs did no alter the growth parameters of the WT yeast strain, not expressing αSyn. As expected, the expression of αSyn drastically decreased the growth of BY4741 cells (Fig. 1A). PEFs enhanced the growth of the yeast cells expressing αSyn, with the exception of A. unedo leaf and C. album fruit PEFs (Fig. 1A). After a longer lag phase, cells exposed to R. idaeus fruit, G. biloba leaf, A. unedo fruit and C. album leaf recovered and achieved growth rates similar to those of the control cells (Fig. 1A). The metabolic capacity assay, based on the ability of living cells to convert resazurin to resorufin, showed that C. album leaf PEF promoted the highest level of protection against αSyn toxicity, followed by C. album fruit, A. unedo fruit and G. biloba leaf (Fig. 1B).

The formation of inclusions is related with αSyn-induced toxicity with concomitant reduction of cellular viability (26,28). To assess if the enhancement in cell viability was followed by alterations in the sub-cellular localization of αSyn, cells expressing αSyn fused with GFP (αSyn-GFP) were treated with PEFs and the percentage of cells with αSyn inclusions was monitored by fluorescence microscopy (Fig. 1C). Interestingly, treatment with C. album leaf PEF reduced the percentage of cells displaying αSyn inclusions. Arbutus unedo leaves also induced a small decrease in αSyn inclusion formation (Fig. 1C).

The effect of the PEFs on oxidative stress was also analyzed by subjecting BY4741 WT, Δyap1, Δsod1 and AD1–8 cells to H₂O₂ (Supplementary Material, Fig. S3). Corema album leaf PEF induced a mild protection in Δsod1 and AD1–8 cells.

Overall, these results suggest that C. album leaves PEF is the strongest cytoprotective agent in the yeast model of PD. Interestingly, the PEFs presented different composition and protective capacities, affecting distinct parameters. The PEF from C. album leaves counteracted the cytotoxicity induced by αSyn expression (Fig. 1A and B), was capable of reducing the percentage of cells displaying αSyn inclusions (Fig. 1C) and enhanced the growth phenotype of yeast strain under H₂O₂-induced toxicity (Supplementary Material, Fig. S3).

Corema album leaf PEF comprises a complex mixture of (poly)phenols (Supplementary Material, Fig. S1 and Table S3). Myricetin-3-O-galactoside (peak 4, Supplementary Material, Fig. S1 and Table S3), myricetin-3-O-glucoside (peak 5, Supplementary Material, Fig. S1 and Table S3) and quercetin-3-O-galactoside (peak 13, Supplementary Material, Fig. S1 and Table S3) are three of the main identified (poly)phenols. In order to verify if these compounds could be responsible for the observed protection against αSyn toxicity, we compared their activity with C. album leaf PEF in the yeast PD model (Fig. 1D). Yeast cells expressing αSyn were treated with a range of concentration from 0.5 to 30 µg ml⁻¹ of the isolated compounds dissolved in the same medium as PEF, and viability was assessed by flow cytometry (FCM) using propidium iodide (PI), a fluorescent molecule that is only absorbed by cells that have lost membrane integrity. At the range of concentrations tested, the pure compounds separately did not show any efficacy in reducing cell death (in comparison to non-treated cells expressing αSyn). Remarkably, C. album leaf PEF caused a dose-dependent reduction of αSyn toxicity (≈40%).

Protection against αSyn cytotoxicity is mediated by the reduction of ROS

To further characterize the effect of C. album leaf PEF on αSyn-induced toxicity and aggregation, yeast cells expressing αSyn-GFP were treated with this PEF and analyzed by FCM for αSyn-GFP and PI fluorescence (Fig. 2A). Cell doublets were excluded from all FCM analyses.

Figure 2.

Corema album leaf PEF reduces αSyn-induced toxicity, aggregation and ROS levels. BY4741 cells were treated with C. album leaf PEF for 6 h in raffinose medium, prior to treatment for 12 h with PEF in galactose medium (αSyn ON). Cells transformed with the αSyn empty vector were use as control. (A) Cell viability and αSyn expression evaluated by PI versus αSyn-GFP fluorescence, respectively, assessed by FCM. (B) Histograms of number of cells versus PI fluorescence intensity assessed by FCM, showing only the PI-positive population (left panel), and the correspondent percentage of PI-positive cells (right panel). (C) Cell viability assessed by spot assays of cells subjected to the treatment. (D) Histogram of the number of cells versus αSyn-GFP MFI, showing only the GFP-positive population (left panel) and the correspondent MFI values (right panel), assessed by FCM. Regions marked represent sub-population 1 and sub-population 2 of cells exhibiting different αSyn-GFP fluorescence intensity. (E) Fluorescence microscopy images of cells subjected to the treatment, with αSyn localized either in the membrane or in intracellular inclusions (cells marked with 1 and 2, respectively). (F) αSyn expression levels of cells subjected to the treatment (GAPDH was used as loading control), assessed by western blot. (G) ROS MFI, assessed by FCM using the DCFHDA probe. (H) Superoxide radical MFI assessed by FCM using the DHE probe. Values represent the mean ± SD of at least three independent experiments. Statistically significant differences between the indicated treatments are shown, *P < 0.05, **P < 0.01, ***P < 0.001 and n.d. as not different for P < 0.05.

αSyn expressing cells presented a higher percentage of PI-positive cells compared with the control where αSyn was not expressed (Fig. 2A and B). This is in agreement with the previous results (Fig. 1A and B) and with the known cytotoxic effects of the human protein (11,28).

When cells expressing αSyn were treated with C. album leaf PEF, an increase in cell viability was observed (Fig. 2A and B). Similarly, this PEF ameliorated the growth phenotype of these cells (Fig. 2C). These observations are in agreement with the detected increase in metabolic capacity of the cells (Fig. 1B).

Using FCM, we observed a reduction in αSyn median fluorescence intensity (MFI) mediated by C. album leaf PEF (Fig. 1D). This was correlated with the ability of this PEF to reduce the percentage of cells with αSyn inclusions by 28.2%, as confirmed by fluorescence microscopy (Fig. 1C and 2E, respectively).

Interestingly, cells expressing αSyn-GFP presented two fluorescence intensity peaks corresponding to two different sub-populations (regions marked as 1 and 2 in Fig. 2D). The higher MFI observed for sub-population 2 is due to the presence of αSyn inclusions with higher fluorescence intensity. In fact, GFP fluorescence intensity was stronger in cells presenting αSyn inclusions (Fig. 2E, cells marked with 2) than in cells where αSyn was localized at the cell membrane (Fig. 2E, cells marked with 1). These different intensities were clearly visible when cells were observed by fluorescence microscopy (Fig. 2E) (26). Corema album PEF reduced the population with αSyn inclusions (sub-population 2) by 25.2% (Fig. 2D). Together with the reduction of αSyn inclusions (Fig. 1C) and αSyn fluorescence intensity (Fig. 2D and E), these data confirm that C. album leaf polyphenols restore the sub-cellular localization of αSyn to the membrane. Next, αSyn expression was analyzed by immunoblot (Fig. 2F). Importantly, we found that these (poly)phenols did not alter the levels of αSyn.

Oxidative stress is a known trigger of αSyn misfolding (12,13). ROS levels were determined by FCM using 2′,7′-dichlorfluorescein-diacetate (DCFHDA) (Fig. 2G), an indicator of the general oxidative state of the cell (41). αSyn expression in yeast induced an increase in ROS accumulation (Fig. 2G) as previously described (42). PEF treatment reduced the basal levels of ROS both in the control and in cells expressing αSyn.

To gain insight into the redox homeostasis status, superoxide radical levels were also evaluated as an indicator of mitochondrial metabolic function (43), since these organelles are damaged by αSyn-induced toxicity (44). Superoxide was analyzed by FCM using a specific probe, dihydroethidium (DHE) (41), and the results indicated that αSyn expression led to an increase in superoxide production (Fig. 2H). Treatment with PEF reduces the basal levels of this radical in control cells as well as in cells expressing αSyn (Fig. 2H). Thus, the protective role of C. album PEF may be related with the improvement of redox homeostasis and mitochondrial function.

(Poly)phenols promote the clearance of αSyn by stimulating autophagy

One of the mechanisms implicated in the pathogenesis of PD is the impairment of the protein clearance pathways (4,45). In yeast, the clearance of αSyn inclusions has been associated with both ubiquitin–proteasome system (UPS) and autophagy–lysosome pathway (46). First, we evaluated αSyn clearance at 24 h by western blotting (Fig. 3A). Interestingly, cells treated with PEF presented lower levels of αSyn at 24 h of clearance, suggesting that the (poly)phenols promoted the degradation of αSyn. Thus, we next asked whether C. album leaf PEF protected against αSyn toxicity by interfering with the activity of αSyn proteostasis pathways.

Figure 3.

Corema album leaf PEF promotes the clearance of αSyn by stimulating autophagy without changing the proteasomal function. BY4741 cells were treated with C. album leaf PEF for 6 h in raffinose medium, prior to treatment for 12 h with PEF in galactose medium (αSyn ON). Cells transformed with the αSyn empty vector were use as control. (A) Clearance of αSyn evaluated by western blotting at the indicated time points (left panel) and respective densitometry normalized to 0 h of clearance (right panel). PGK was used for normalization. (B) Proteasome impairment evaluated by FCM using GFPu MFI. (C) Autophagy evaluated by GFP-Atg8 processing assay assessed by western blot (left panel). ATG8 induction quantified by the total GFP signal (GFP-Atg8 and free GFP signal, detected with anti-GFP) (middle panel); autophagic flux quantified by measuring the vacuolar degradation of the Atg8 domain reporter (ratio of free GFP to total GFP signal) (right panel). Rapamycin was used as a positive control. (D) Autophagy and αSyn expression evaluated by mCherry-Atg8 versus αSyn-GFP fluorescence, respectively, assessed by FCM (left panel); corresponding mCherry-Atg8 MFI (middle panel); and percentage of cells αSyn-GFP and mCherry-Atg8 positive, determined only in the αSyn-GFP-positive population of cells (right panel). Values represent the mean ± SD of at least three independent experiments. In the upper panel, the staining controls are represented. Statistically significant differences between the indicated treatments are shown, *P < 0.05, **P < 0.01, ***P < 0.001 and n.d. as not different for P < 0.05.

The effect of PEF on the proteasome activity was analyzed using an unstable GFP reporter (GFPu) consisting of a fusion of GFP with a constitutive degradation signal (CL-1) that promotes its rapid degradation by the UPS (28). FCM analysis was performed in cells co-expressing αSyn and GFPu and we found that upon αSyn expression the levels of GFPu levels increased, as previously reported (Fig. 3B) (28). Treatment with C. album leaf PEF did not affect GFPu levels, independently of αSyn expression (Fig. 3B).

Subsequently, we tested whether autophagy was affected using GFP-Atg8 as a reporter (47,48). Atg8 is one of the key molecules involved in autophagy. Its conjugation to the autophagosomal membrane, through an ubiquitin-like conjugation system, is essential for autophagy in eukaryotes (48). Autophagy is induced by the inhibition of the Tor kinase with rapamycin, thus it was used as a positive control. The GFP-Atg8 processing assay showed that αSyn expression leads to a significant reduction on both ATG8 induction and autophagic flux (Fig. 3C) suggesting an interference in autophagy function. Interestingly, when αSyn expressing cells were treated with C. album leaf PEF, the autophagic flux was restored to control levels (Fig. 3C). Although we observe a tendency to increased autophagic induction and flux in control cells exposed to PEF in the tested conditions, this effect is not statistical significant (Supplementary Material, Fig. S4). Autophagy was also monitored by FCM using mCherry-Atg8, as mCherry fluorescence is more stable than GFP at vacuolar pH (48) (Fig. 3D). We observed a reduction in mCherry-Atg8 MFI upon αSyn expression, and an increase in cells treated with rapamycin (Fig. 3D, middle panel), in agreement with the GFP-Atg8 processing assay (Fig. 3C). Interestingly, with this assay we also detected an increase in mCherry-Atg8 MFI in cells expressing αSyn treated with PEF (Fig. 3D, middle panel), and an increase in the percentage of cells αSyn-GFP and mCherry-Atg8 positive (Fig. 3D, right panel).

By analyzing αSyn expressing cells using FCM, we observed difference in ATG8 induction that was not detectable in the GFP-Atg8 processing assay (Fig. 3C), mainly due to the possibility of analyzing only the population of cells expressing αSyn. These observations indicate that (poly)phenols stimulate autophagy in yeast cells expressing αSyn (Fig. 3C and D).

(Poly)phenols inhibit αSyn fibrillization in vitro and promote the formation of stable, non-toxic oligomeric species

In order to assess if the protection observed was due to a direct modulation of αSyn aggregation, we investigated the fibrillization of recombinant αSyn in the presence of 0.5 or 30 µg GAE ml⁻¹ of C. album leaf PEF (Fig. 4A). The kinetics of αSyn fibril formation was monitored using thioflavin T (ThT) as previously described (49). Upon αSyn fibrillization, we observed the expected increase in ThT fluorescence signal, confirming the formation of amyloid-like fibrils. When recombinant αSyn was incubated with C. album leaf PEF, no significant increase in ThT fluorescence was observed (more evident in the presence of 30 µg GAE ml⁻¹), suggesting that fibril formation was inhibited.

Figure 4.

Corema album leaf PEF reduces αSyn fibrillization and promotes the formation of non-toxic αSyn species. αSyn (70 µm) was allowed to fibrillization in the presence of C. album leaf PEF, and samples were collected over time. (A) Fibril formation was monitored by the increase in ThT fluorescence. αSyn alone (filled circle), with 0.5 (gray filled triangle) or 30 (open square) µg GAE ml⁻¹C. album leaf PEF. (B) Western blot of αSyn species separated by SDS–PAGE. (C) Effect of αSyn species in H4 viability, assessed by LDH release. H4 cells were treated for 6 h with αSyn resulting species from vehicle, αSyn 0.5 or 30 µg GAE ml⁻¹C. album leaf PEF treatment obtained before (0 h) and after (48 h) fibrillization. Cells treated with vehicle or PEF alone were used as controls. Results are normalized to the vehicle control cells. (D) Triton X-100 soluble (T-Soluble) and Triton X-100 insoluble (T-Insoluble) fractions of yeast cells expressing αSyn treated with PEF, assessed by western blotting (upper panel) and determination by densitometry analysis of the percentage of T-Insoluble αSyn (lower panel). Values are normalized to the total fraction. Values represent the mean ± SD of at least three independent experiments. Statistically significant differences between the indicated treatments are shown, *P < 0.05, **P < 0.01, ***P < 0.001 and n.d. as not different for P < 0.05.

The biochemical nature of the αSyn species was further evaluated by SDS–PAGE (Fig. 4B). We observed that C. album leaf PEF promoted the formation of SDS-stable, high-molecular oligomeric αSyn species, an effect that was previously described with pure (poly)phenols (50,51).

To evaluate the toxicity of these oligomeric αSyn species obtained after C. album leaf PEF incubation, human neuroglioma H4 cells were treated with 1 µm of αSyn species and LDH activity was assessed in the medium as a measure of membrane integrity (Fig. 4C). We compared the toxicity of αSyn species before (0 h) and after (48 h) fibrillization in the presence/absence of PEF. Cells treated with the vehicle or PEF alone were used as controls. While the αSyn species formed in the absence of PEF were toxic, those formed in the presence of C. album leaf PEF were not (Fig. 4C). In agreement, a previous report showed that stable oligomeric αSyn species formed in the presence of EGCG were not toxic (50).

To gain further insight into the effect of PEF on αSyn, we analyzed the Triton X-100 detergent solubility of the αSyn species formed in yeast cells (Fig. 4D). The treatment with PEF resulted in a reduction of αSyn in the soluble fraction and consequent increase in the insoluble fraction. These results are consistent with the in vitro ThT assay, suggesting that (poly)phenols promoted the stabilization of oligomeric species both in vitro and in yeast cells.

(Poly)phenols protect human cells from H₂O₂-induced toxicity, modulate αSyn aggregation and promote autophagy

The protective capacity of C. album leaf PEF was studied in human H4 cells treated with H₂O₂ (Fig. 5A). The pretreatment with PEF protected cells from oxidative stress-induced cell death, as observed by the reduction of PI-positive cells (Fig. 5A). These data show that the PEF protects human cells from oxidative stress, in line with the observed reduction of ROS and superoxide radical in yeast (Fig. 2G and H).

Figure 5.

Corema album leaf PEF protects against H₂O₂-induced toxicity, modifies the distribution of αSyn inclusions and induces autophagy in H4 cells. (A) H4 cells were pre-treated with C. album leaf PEF for 16 h, then subjected to 600 µm of H₂O₂ for 6 h. Viability was assessed by PI fluorescence versus side scatter (SSC) (left panel); correspondent percentage of PI-positive cells (right panel), determined by FCM. (B) H4 cells were transfected with SynT and Synphilin-1-V5 (SynT) for 24 h and treated with PEF for 6 h, subsequently the percentage of cells with inclusions was determined by fluorescence microscopy (left panel). Cells were classified by cells without inclusions (open square), with <10 inclusions (gray filled square) or with 10 or more inclusions (filled square) (right panel). (C) αSyn levels of H4 cells transfected with empty vector (Control), co-transfected with SynT and Synphilin-1-V5 (SynT) or treated with 30 µg GAE ml⁻¹C. album leaf PEF (SynT + PEF) cells, assessed by western blot (upper panel). GAPDH was used as loading control. Corresponding densitometry is presented (lower panel). (D) Autophagy evaluated by LC3-II levels in PEF-treated cells. Control cells (not expressing αSyn), cells expressing SynT and Synphilin-1-V5 (SynT) or cells expressing αSyn WT were treated with 20 mm NH₄Cl and 200 µm leupeptin (lysosomes inhibitors) for 2 and 4 h. LC3 levels were determined by western blot (β-actin was used as loading control) (left panel). The basal levels of autophagy were measured (middle panel) and the autophagic flux was given by the difference between LC3-II levels in cells treated or not with lysosomes inhibitors for 4 h (right panel). Results are presented as the ratio of PEF- and vehicle-treated cells. Values represent the mean ± SD of at least three independent experiments. Statistically significant differences between the indicated treatments are shown, *P < 0.05, **P < 0.01, ***P < 0.001 or n.d. as not different for P < 0.05.

When studying αSyn aggregation, we observed that the number of αSyn inclusions per cell was modulated by the PEF (Fig. 5B). Specifically, PEF treatment increases the percentage of cells presenting >10 inclusions. Importantly, the effect was independent of αSyn levels (Fig. 5C).

Subsequently, we evaluated autophagy by measuring LC3-II levels in human H4 cells (Fig. 5D, left panel), LC3-II is the mammalian homolog of Atg8 and is used as an autophagosomal marker (48,52). In cells expressing αSynWT treated with PEF, we observed a significant increase in the LC3-II basal levels (Fig. 5D, middle panel). Furthermore, the autophagic flux was evaluated by determining the accumulation of LC3-II upon impairment of autophagy with lysosomal inhibitors (48,52). PEF treatment increased the autophagic flux in both αSynWT and SynT cells (right panel). In agreement with the observations in yeast, these results also suggest that (poly)phenols promote autophagy in αSyn expressing cells (Fig. 3).

Discussion

PD and other synucleinopathies are currently incurable disorders affecting a growing number of individuals, due to the aging of the human population. Thus, there is an urgent demand for the identification of novel strategies for therapeutic intervention. Oxidative stress has long been implicated in PD and other neurodegenerative diseases, constituting an attractive pathway for intervention. Given that public health policies are currently focused on the prevention of cognitive decline, through the promotion of physical exercise, cognitive and social activity, and natural nutrition and supplementation, (poly)phenols play an attractive role as they can easily be incorporated as dietary supplements (16,17).

In this study, PEFs from A. unedo, C. album, G. biloba and R. idaeus obtained from leaves and/or fruits were analyzed by LC–MS and evaluated for in vitro antioxidant capacity. Furthermore, the cytoprotective activities of these PEFs were analyzed using established yeast and human cell models of PD/synucleinopathies (23,28).

Phenolic compounds, pure or in extracts, are of great interest in nutrition and medicine because of their beneficial effects on human health, demonstrated in both pre-clinical and clinical trials (16,19,32–34). However, a detailed understanding of the chemical origin underlying the bioactivities of these plant species is essential for the rational use of these compounds. Berries are known for their high content in (poly)phenols and in particular anthocyanins, usually associated to their high antioxidant properties (53). Interestingly, the berries analyzed in this study have distinct levels of this phenolic class that do not reflect on their in vitro antioxidant activity. In particular, PEFs from A. unedo and C. album fruits, with lower anthocyanin content, exhibited higher antioxidant capacity than R. idaeus, a species rich in anthocyanins. We identified other phenols, like gallic acid, chlorogenic acid derivatives and flavonols as contributors to the antioxidant capacity observed for that species.

With respect to the leaves, C. album PEF, characterized here for the first time, presented the highest antioxidant capacity. The main difference between this PEF and the ones from A. unedo and G. biloba leaves is the content in myricetin glucosides and stilbene derivatives. Stilbenes have emerged as promising molecules in human health, being resveratrol the most studied one (54).

Saccharomyces cerevisiae was used as an eukaryotic cell model to study the bioactivity of PEFs. By using strains with distinct genetic backgrounds that resulted in different susceptibilities to oxidative stress and compound permeability (31,38), we were able to conclude that a higher in vitro antioxidant capacity was not directly related to an effective in vivo antioxidant protection of the studied PEFs. This observation reinforces the idea that other mechanisms, in addition to radical scavenging activity, may be involved in the protection by (poly)phenols.

While the majority of PEFs counter balanced the absence of SOD1, only the leaf and fruit (poly)phenols from A. unedo were able to protect yeast cells upon YAP1 deletion. This gene encodes a transcription factor that plays an important role in the oxidative stress response (55,56). These findings suggest a specific interaction between A. unedo (poly)phenols and the genes/enzymes regulated by Yap1p should be occurring, in order to allow cells to overcome its absence. Importantly, some flavonoids modulate the expression of known targets of Yap1p, namely γ-glutamylcysteine synthetase (Gsh1p) (57).

The diverse chemical composition of the PEFs and the distinct genetic background of the yeast strains tested might account for their different bioactivities. In fact, the cellular absorption of (poly)phenols may differ greatly and is influenced by factors such as chemical structure, which affects their biological properties, antioxidant activity, specific interactions with cell receptors and enzymes (58). For instance, aglycone forms of (poly)phenols are more hydrophobic than its glycosylated forms which are highly hydrophilic. Previous studies showed that aglycone forms may have the ability to cross cellular membranes through diffusion, while the absorption of the (poly)phenols glucosides is thought to occur via transporters in mammalian cells (58–60). In the PEFs tested in this study, the main compounds are glycosylated, except in C. album fruit that is characterized by the presence of non-glycosylated phenolic acids, although conjugated to quinic acid (37).

To further characterize the bioactivities of the PEFs, we tested them in both yeast and human cell models of PD (24,28). The cytotoxicity induced by αSyn expression was evaluated by complementary approaches, with respect to cell viability, αSyn inclusion formation and ROS production.

Our data demonstrate that C. album leaf PEF was the most promising protective agent against both oxidative stress conditions and αSyn-induced toxicity. To investigate the mechanism of action of C. album leaf PEF, we explored its effect on different cellular pathways related to αSyn toxicity (summarized in Fig. 6).

Figure 6.

Putative mechanism of action of C. album leaf PEF on αSyn cytotoxicity. αSyn misfolding and aggregation promotes toxicity and cell death through disruption of several cellular functions and by increasing ROS. Mitochondria, one of the main organelles damaged, work both as a generator and a target of ROS. In turn, oxidative stress can promote αSyn aggregation by directly oxidizing it and/or indirectly by perturbing its clearance, generating a vicious cycle. Healthy cells are able to counteract αSyn-induced toxicity and aggregation by its clearance through autophagy and UPS, pathways known to be affected in pathological conditions. Corema album leaf PEF protection could occur through (poly)phenols interaction with αSyn, inhibiting its fibrillization, and promoting the formation of stable non-toxic (protective) species, a feature of (poly)phenol observed in vitro. Indirectly, C. album leaf PEF ameliorate the general redox state of the cell, by acting as ROS scavenger and/or by promoting mitochondria function. In parallel, C. album leaf PEF promotes the clearance of αSyn and restores a functional autophagic flux, disrupted by αSyn toxicity. Overall, C. album leaf PEF modulates αSyn toxicity, leading to an increase in cell viability.

Corema album leaf PEF enhances the viability of yeast cells while reducing the percentage of cells displaying αSyn inclusions and promoting its clearance. In human H4 cells, the same PEF protects from H₂O₂-induced toxicity and modulates the distribution of αSyn inclusions, increasing the percentage of cells with >10 inclusions. Our studies using recombinant αSyn revealed that C. album leaf PEF directly interferes with fibrillization and promotes the formation of stable high-molecular weight αSyn species. Interestingly, these species are less toxic when added exogenously to cells. The in vitro fibrillization observations go in line with the increased insolubility of αSyn species promoted by PEF, in the yeast PD model. Thus, by stabilizing these non-toxic species, the PEF may be reducing the formation of more toxic species (10,51). In fact, previous studies investigated the direct effect of isolated (poly)phenols on amyloid fibril formation (22,49,61). Myricetin and quercetin were described to efficiently inhibit αSyn fibrillization (10). In C. album leaves, the main compounds identified were myricetin-3-O-glucoside, myricetin-3-O-galactoide and quercetin-3-O-glucoside. We compared these isolated compounds with C. album leaf PEF for their capacity to protect against αSyn-induced toxicity. The isolated compounds did not reproduce the PEF effect, suggesting that synergistic effects between PEF compounds are necessary and/or that other less-abundant compounds contribute to the overall protective effect. In fact, in line with the hypothesis of synergistic effects, previous studies showed that total extract from G. biloba (known for its medical applications) is more active than its isolated (poly)phenols (32,62). Therefore, it is reasonable to speculate that the therapeutic benefits of (poly)phenol extracts depend on the combination of its components, and on their synergistic action on diverse cellular processes. This is a rather important feature of (poly)phenol extracts in the context of multifactorial disorders, caused by a combination of genes and environmental factors and comprising the failure/impairment of several cellular pathways, such as PD.

Corema album leaf PEF decreased αSyn inclusion formation by directly interacting with αSyn, stabilizing non-toxic oligomers or by indirect mechanisms, such as through the reduction of oxidative stress and/or the stimulation of αSyn degradation pathways.

Regarding oxidative stress, we found that C. album leaf PEF promoted a general reduction in the levels of ROS and in particular of superoxide radical in yeast. Superoxide is the major ROS produced by the mitochondria (50,51). Therefore, the reduction in superoxide might be related with a more direct role of the PEF in the rescue of mitochondrial dysfunction induced by αSyn toxicity (44).

αSyn degradation occurs both via the UPS and the autophagy–lysosome pathway (as reviewed in (46)). The UPS is thought to be primarily responsible for the degradation of short-lived signaling molecules or proteins tagged with ubiquitin (46). Interestingly, overexpression of αSyn causes a decrease in proteasome function (28,63). However, the effects of proteasomal inhibition on αSyn degradation are still controversial. While some studies reported that αSyn appeared not to be degraded by UPS (8), others reported that proteasomal inhibition triggered accumulation and aggregation of αSyn (64). The relevance of proteasome activity in PD is highlighted by reports that describe a decrease in the proteasome peptidase activity in the substantia nigra of sporadic PD patients, accompanied with a decreased expression of proteasomal α-subunits in the same brain region (4,65). In the conditions tested in our study, αSyn expression also led to proteasome impairment, as described before (28,63). Nevertheless, our results show that the mechanism responsible for the PEF protection, of αSyn cytotoxicity, is not directly related with modulation of proteasome function in the yeast PD model.

Under stress conditions, the UPS can be overwhelmed, and the autophagy pathway is then required to compensate for the increased protein accumulation and potential damage. Thus, we evaluated autophagy as a possible target of PEF bioactivity. αSyn expression inhibited the autophagic induction and flux in yeast. Recent findings relate autophagy dysfunction to αSyn-induced toxicity in a variety of PD models, from yeast to higher organisms (6,43,45,66–69). Remarkably, upon incubation with C. album leaf PEF, the autophagy flux impaired by αSyn expression was restored. These results, together with the observation that PEF incubation reduces the percentage of cells with αSyn inclusions and promotes its clearance, suggest that the protection promoted by this PEF may be related with increased degradation of αSyn by autophagy. In particular, recent studies indicate that the clearance of αSyn inclusions is more dependent on the autophagic pathway than on the UPS (5,6). We confirmed these observations in a human cell model, where we found that C. album leaf PEF exhibited cytoprotective properties by modulating the distribution of αSyn inclusions, protecting cells against oxidative injury induced by H₂O₂ and by promoting autophagy of αSyn expressing cells.

Our study identifies C. album leaf PEF as very promising agents that rescue the autophagic flux and/or decrease αSyn fibrillization resulting in the formation of non-toxic species. Our findings also suggest that PEF may increase the accessibility of αSyn inclusions to autophagic degradation. In total, our findings open a new window of opportunities for drug development, as there is accumulating evidence supporting the relevance of autophagy and αSyn modulation as a therapeutic targets in PD and other synucleinopathies (6,10,43,51,69).

Materials and Methods

Plant material and extraction procedure

Fruits and leaves of A. unedo L. and C. album L. were collected by random sampling in an extensive area of Arrábida Natural Park and Comporta (southern region of Portugal), respectively. Rubus idaeus cv. Polka was grown in Fataca experimental field, Odemira, Portugal and G. biloba was grown in the Institute green house. Plant material was extracted as described in (70).

Extract fractionation by solid phase extraction

Hydroethanolic extracts were performed and fractionated by solid phase extraction as described in (38).

Total phenolic quantification

Determination of total phenolic compounds was performed by the Folin-Ciocalteau method adapted to microplate reader (71).

Peroxyl radical scavenging capacity determination

Peroxyl radical scavenging capacity was determined by the oxygen radical absorbance capacity method as described (31).

Phenolic profile determination by LC–MS

(Poly)phenol fractions were applied to a C-18 column (Synergi Hydro C18 column with polar end capping, 4.6 × 150 mm, Phenomenex^®) and analyzed on an LCQ-DECA system controlled by the XCALIBUR software (2.0, ThermoFinnigan), as described (31).

Yeast strains, transformation and plasmids

Four strains of S. cerevisiae were used in this study. BY4741 (MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0) and its isogenic mutants Δsod1 (MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0; Δsod1::KanMX4) and Δyap1 (MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0 Δyap1::KanMX4) were acquired from Euroscarf (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/) and carry a KanMx4 deletion cassette replacing the gene in question. The AD1-8 strain has eight deletions of so-called multidrug resistance genes (Δyor1, Δsnq2, pdr5-Δ2, Δpdr10, Δpdr11, Δycf1, pdr3-Δ2 and Δpdr15) encoding for ABC transporters involved in xenobiotics efflux (72). Yeast cells were transformed with the indicated plasmids using lithium acetate standard method (73). The empty plasmids p426GAL, p413GPD and pRS415 were used as control (74). The p316-GFP-ATG8 plasmid (47) was a kind gift from Prof Yoshinori Oshumi (National Institute for Basic Biology, Okazaki, Japan) and was used to sub-clone GFP-ATG8 as well as the endogenous ATG8 promoter into the SacI—XhoI sites of pRS415 plasmid (74). Other plasmids used in this study were previously described: p425GAL-αSyn-GFP, p426GAL-αSyn-GFP and p426GAL-αSyn plasmids carrying the human gene of αSyn with or without a C-terminal fusion to GFP under the regulation of GAL1 inducible promoter (28); p316-2xmCherry-ATG8, expressing Atg8 with an N-terminal fusion to mCherry under the regulation of the endogenous ATG8 promoter (a kind gift from Dr. Kuninori Suzuki, Tokyo Institute of Technology) (75); p413GPD-GFPu expressing an unstable GFP under the regulation of GPD constitutive promoter.

Yeast growth conditions and compound testing

A pre-inoculum was prepared until the optical density at 600 nm (OD_{600 nm}) reached ∼0.5 ± 0.1 (log growth phase) in synthetic complete (SC) medium [0.67% (w/v) yeast nitrogen base without amino acids (Difco), 2% (w/v) glucose and 0.79 g l⁻¹ complete supplement mixture (QBiogene)], under shaking at 30°C. Afterwards, cells were diluted in the same medium to obtain all strains in log growth phase. Readings were performed in a 96-well microtiter plate using a plate spectrophotometer (Biotek Power Wave XS). To determine the lag and doubling time the strains were grown in SC medium, as described before, and diluted to an OD_{600 nm} of 0.03 in the same medium supplemented with 250, 500 or 1000 µg GAE ml⁻¹ of the indicated PEF in a 96-well microtiter plate. The cells were then incubated at 30°C with shaking for 24 h. Yeast growth was kinetically monitored hourly during 24 h by measuring the OD_{600 nm} and the growth curves were obtained. For each treatment, doubling and lag phase times were determined based on the linear regression of the log growth phase equation, obtained after logarithmic transformation of OD_{600 nm} values.

For spot assays with H₂O₂, the strains were grown as described and afterwards were inoculated (OD_{600 nm} 0.03) in SC medium with 1000 µg GAE ml⁻¹ of PEFs. After 6 h of growth, cells were pelleted and resuspended in SC medium supplemented with 2 mm H₂O₂. Subsequently, 30 µl of cells were withdrawn and OD_{600 nm} was adjusted to 0.05 ± 0.005 with SC medium. Serial dilutions were made with a ratio of 1 : 4 and 4 µl of each dilution was spotted in YPD solid medium and grown for 2 days at 30°C.

For growth curves with yeast cells expressing αSyn, cells were grown in the appropriated SC selective liquid medium containing 1% (w/v) raffinose, until log phase was reached. Subsequently, cells were incubated for 6 h at 30°C with shaking on the same medium not supplemented or with PEFs. Afterwards, cells were centrifuged and diluted to OD_{600 nm} 0.03 in new SC selective medium 2% (w/v) glucose (control, αSyn OFF) or 1% (w/v) galactose (αSyn ON) not supplemented or with PEFs, and grown for 24 h at 30°C with shaking. The growth was kinetically monitored each hour by measuring OD_{600 nm}.

For αSyn cytotoxicity evaluation, spot assays were performed on solid medium using cells treated with PEFs as described above. OD_{600 nm} was set to 0.05 ± 0.005 and 1 : 4 serial dilutions of each sample were prepared. Then, 4 µl of each dilution was spotted in solid SC selective medium containing 2% (w/v) glucose (αSyn OFF) and incubated at 30°C for 42 h. Images were acquired using Chemidoc^™ XRS and Quantity-one^® software. To evaluate cells viability by FCM with PI, cells were subjected to the previously described treatment with C. album leaf PEF or the pure compounds, myricetin-3-O-galactoside, myricetin-3-O-glucoside or quercetin-3-O-galactoside (Extrasynthese, Genay Cedex, France), in a range of concentration from 0.5 to 30 µg GAE ml⁻¹. Data analysis was performed using FlowJo software and the cell doublets exclusion was performed based on Forward-A and -W scatter parameters. A minimum of 10 000 events were collected for each experiment.

For αSyn clearance experiments, after 12 h of αSyn expression induction the cells were centrifuged washed in PBS, resuspended in 2% (w/v) glucose SC liquid media (αSyn expression OFF) and incubated at 30°C, with shaking, for 24 h. The levels of αSyn were determined by western blotting at 12 h of induction (corresponding to 0 h of clearance) and at 24 h of clearance.

Yeast metabolic capacity assay

Yeast cells were grown as described above. The growth was kinetically monitored each hour by measuring OD_{600 nm} and metabolic capacity was assessed at 6 h of growth using CellTiter-Blue^® Cell Viability assay (Promega), following manufacturer instructions.

H4 cell culture and transfections

Human H4 neuroglioma cells (gift from Dr. Bradley T. Hyman, Harvard Medical School) were maintained at 37°C in OPTI-MEM I (Gibco, Invitrogen, Barcelona, Spain) supplemented with 10% (v/v) fetal bovine serum and seeded at 80 000 cells cm² density 24 h prior to transfection. Cells were transfected as previously described (76) with pcDNA3.1-αSynT and pcDNA3.1-Synphilin-1-V5.

Compound testing in H4

For H₂O₂ toxicity, assay cells were treated with 30 µg GAE ml⁻¹C. album leaf PEF for 16 h, the medium was removed and cells were treated with 600 µm H₂O₂ for 6 h. To evaluate the effect of PEFs in αSyn aggregation, 24 h posttransfection cells were treated with 30 µg GAE ml⁻¹C. album leaf PEF for 6 h.

Western blotting

For αSyn quantification, total yeast protein extraction was performed by glass bead lysis as described before (77,78). For GFP-Atg8 quantification, protein extraction was performed using the method described by Cheong et al. (79). For GFP-Atg8 quantification, cells were treated with 10% (v/v) trichloroacetic acid, and washed twice with acetone. The dry cell pellet was then resuspended in MURB buffer (50 mm Na₂HPO₄, 25 mm MES, pH 7.0, 1% (v/v) SDS, 3 m urea, 0.5% (v/v) 2-mercaptoethanol, 1 mm NaN₃ and 0.05% (w/v) bromophenol blue) with proteases and phosphatases inhibitors (Roche, Mannheim, Germany), and disrupted by vortex with an equal volume of acid-washed glass beads, for 5 min. The samples were incubated at 70°C for 10 min and cell debris were removed by centrifugation. After extraction, western blot was performed following standard procedures.

Using the GFP-Atg8 processing assay, ATG8 induction and autophagic flux were determined (80). ATG8 induction was quantified by the determination of the fold increase of total GFP signal (GFP-Atg8 and free GFP signal, detected with anti-GFP) normalized to PGK; autophagic flux was quantified by measuring the vacuolar degradation of the Atg8 domain reporter (ratio of free GFP to total GFP signal) (81).

For αSyn quantification in human H4 neuroglioma cells were lysed with NP-40 lysis buffer in the presence of protease and phosphatase inhibitor cocktail (Roche, Mannheim, Germany), western blot was performed following standard procedures. Autophagy was quantified as the accumulation of the autophagosome-associated protein LC3. Cells were treated with 20 mm NH₄Cl and 200 µm leupeptin for 2 or 4 h, followed by immunoblotting for LC3. Flux was calculated as the ratio of treated (4 h) to untreated samples after normalization to the loading control, adapted from (48,52). Basal autophagy and autophagic flux results are presented as the ratio of PEF- and vehicle-treated cells.

Antibodies used: αSyn (BD Transduction Laboratories, San Jose, CA, USA), GAPDH (Ambion, Cambridgeshire, UK), GFP (Antibodies Incorporated, Davis, CA, USA), PGK (Life Technologies, Paisley, UK), LC3 (nanoTools, Teningen, Germany) and β-actin (Abcam, Cambridge, UK).

Flow cytometry

FCM was performed in a FACS BD LSR Fortessa, equipped with the 695/40 BP and the 685 LP. To determine the levels of ROS, yeast cells were incubated with 50 µm DCFHDA (Molecular Probe, Life Technologies) or 30 µm DHE (Molecular Probe, Life Technologies), for 15 min at 30°C, with agitation and protected from light (41). To analyze cell viability with PI, yeast or H4 cells were incubated with 5 µg ml⁻¹ of PI, for 30 min protected from light. To study the proteasome impairment, yeast cells were transformed with the αSyn and GFPu encoding plasmid or the respective empty plasmid. Autophagy was analyzed in cells transformed with the mCherry-Atg8 and αSyn-GFP encoding plasmids or with the empty plasmids in a BD FACSAria, cells treated with rapamycin were used as control. Data analysis was performed using FlowJo software and the exclusion of cell doublets was performed based on Forward-A and -W scatter parameters. A minimum of 10 000 events were collected for each experiment. Results were expressed as MFI of a molecule.

Fluorescence microscopy

In order to determine the percentage of yeast cells with αSyn inclusions, cells were grown as described above and GFP fluorescence was visualized using a Leica DM3000 fluorescence microscope. The proportion of cells presenting αSyn inclusions was then determined by counting at least 800 cells for each treatment using ImageJ software. Transfected H4 cells were fixed and permeabilized with methanol and blocked in 1.5% (v/v) normal goat serum in PBS for 1 h. Cells were incubated with primary antibody overnight at 4°C (mouse anti-αSyn; BD Transduction Laboratories, San Jose, CA, USA) followed by secondary antibody incubation for 1 h (goat anti-mouse IgG-Alexa488, Invitrogen Corporation, Carlsbad, CA, USA). Slides were subjected to fluorescence microscopy with a Zeiss Axiovert 200 M Widefield Fluorescence microscope. The proportion of cells with αSyn inclusions within the population was then determined by counting at least 100 cells per condition using ImageJ software (25).

Thioflavin T assay

The expression and purification of human αSyn was performed as described in (76). For the ThT assay, the protein solution (70 μm) was mixed with C. album leaf PEF to a final concentration of 0.5 or 30 µg GAE ml⁻¹ in 50 mm Tris–HCl buffer (pH 7.4). Protein samples were stirred at 900 rpm, 37°C in a thermomixer. 1.4 μm of αSyn was added to 1 μm of ThT solution in 50 mm Tris–HCl buffer and ThT fluorescence was recorded at the tested time points as in (82). The samples obtained in the different time points were analyzed by SDS–PAGE using standard western blotting procedures as described above. The toxicity of αSyn species formed at 0 and 48 h of fibrillization was evaluated as described in (83). Cells treated only with the correspondent concentration of C. album leaf PEF were used as controls. Cells treated with 10% (v/v) Triton X-100 for 10 min were used for normalization.

Triton soluble and insoluble fractions

Total protein was extracted and quantified with the BCA protein assay kit (Thermo Fisher Scientific Inc., IL, USA). Total protein (200 µg) was incubated with 1% Triton X-100 on ice, for 30 min. Protein fractions were separated by centrifugation at 15 000g for 60 min at 4°C. The top soluble protein fraction (T-Soluble) was collected and the insoluble protein fraction (T-Insoluble) pellet was resuspended in 40 µl of 2% SDS Tris–HCl buffer, pH 7.4, by pipetting and subsequent sonication (10 s). Total protein, T-Soluble and T-Insoluble fractions (5 µl of each) were loaded and resolved by SDS–PAGE, western blotting was performed as previously described.

Statistical analysis

The results reported in this work are the average of at least three independent biological replicates and are represented as the mean ± SD. Differences among treatments were assessed by analysis of variance with Tukey's honestly significant difference multiple comparison test (α = 0.05) using SigmaStat 3.10.

Funding

This work was supported by Fundação para a Ciência e Tecnologia (Pest-OE/EQB/LA0004/2011, PTDC/BIA-BCM/111617/2009, SFRH/BD/73429/2010 to D.M., IF/01097/2013 to C.N.S., SFRH/BPD/84336/2012 to L.T., SFRH/BPD/35767/2007 to S.T., SFRH/BPD/64702/2009 to H.V.M.). The Scottish Government Rural and Environment Science and Analytical Services Division (D.S. and G.J.M.), Climafruit (Interreg IVB) (D.S.), EUBerry FP7 KBBE-2010-4 265942 (C.N.S. and D.S.) and BacHBerry FP7-KBBE-2013-613793 (C.N.S. and D.S.) Marie Curie International Reintegration Grant (T.F.O.) and an EMBO Installation Grant (T.F.O.). T.F.O. is supported by the DFG Center for Nanoscale Microscopy and Molecular Physiology of the Brain.

Acknowledgements

We acknowledge Márcia Oliveira and Francisco Javier Enguita for the purification of αSyn and for assisting with αSyn fibrillization studies. Prof. Yoshinori Oshumi (National Institute for Basic Biology, Okazaki, Japan) for the p316 GFP-ATG8 plasmid; Prof. Kuninori Suzuki (Tokyo Institute of Technology, Yokohama, Japan) for the 2xmCherry-ATG8 plasmid. Pedro Oliveira (Instituto Nacional de Investigação Agrária, Oeiras, Portugal) for R. idaeus fruits. Regina Menezes and Madalena Reimão-Pinto for the critical review of the manuscript.

Conflict of Interest statement. None declared.