Inhibition and disaggregation of α-synuclein oligomers by natural polyphenolic compounds
Abstract
Alpha-synuclein binds to Alpha-synuclein by biophysical (View Interaction 1, 2) Convergent genetic, biochemical and animal studies over the past decade strongly suggest that the aggregation of the 140-residue pre-synaptic alpha-synuclein (αS) protein plays a fundamental role in the etiology and pathogenesis of Parkinson's disease (PD) and related disorders [1]. These disorders are in fact collectively known as “α-synucleinopathies”, with PD being the most common movement disorder [2]. Motor symptoms in PD reflect the severe degeneration of dopaminergic neurons in the substantia nigra pars compacta. The critical etiological link between αS aggregation and toxicity to dopaminergic neurons was recently demonstrated in vivo in a transgenic Drosophila model of the disorder [3]. Pathogenesis of PD thus appears to be linked to conditions that increase the propensity for αS to aggregate and form fibrils [4]; among these: an increase in the intracellular concentration of αS (e.g. by increased gene copy number) [5, 6], missense mutations [7], oxidative modifications [8, 9], phosphorylation [10], the presence of metal ions (e.g. Fe3+) [11], and interaction with phospholipid membranes [12]. A lot of attention is therefore being directed at the development of molecular inhibitors of αS aggregation for the prevention and treatment of PD [13]. Indeed, an important class of compounds found to be protective against αS fibrillation is that of the polyphenols, which are important beneficial constituents in human diet and medicinal plants [14]. Polyphenols are abundant in a wide range of fruits, vegetables and beverages, including tea, red wine, apples, berries, and strawberries. The daily average polyphenol intake is difficult to estimate, but is supposed to average 200–500 mg per day [14]. Polyphenolic compounds identified with anti-fibrillogenic and fibril-destabilizing effects for αS include baicalein (Baic) [15], epigallocatechin gallate (EGCG) [16, 17], rosmarinic acid (RA) [18], tannic acid (TA), nordihydroguaiaretic acid (NDGA), curcumin, myricetin (Myr), kaempferol, catechin, and epi-catechin [19]. Additional studies showed that several polyphenols inhibited αS filament assembly by forming soluble, non-cytotoxic, oligomeric complexes with the αS protein [16, 20-22]. Even more recent findings suggest that EGCG is a powerful remodelling agent of mature αS-synuclein fibrils, converting them into non-toxic, smaller, amorphous aggregates [17]. In this study, we wanted to make use of the structural diversity of natural polyphenols to define key molecular scaffolds most effective in inhibiting oligomer formation by alpha-synuclein and/or disaggregating pre-formed oligomers. Fourteen polyphenolic compounds and black tea extract (BTE) (Table 1 ) were systematically tested using confocal single-particle fluorescence techniques. In recent years, fluorescence correlation spectroscopy (FCS), fluorescence-intensity distribution analysis (FIDA) and scanning for intensely fluorescent targets (SIFT) have been recognised as powerful tools for highly sensitive analysis of aggregation processes in neurodegenerative diseases, including detection and characterization of synuclein oligomers [23-25]. Importantly, this technique allowed us to use nanomolar and low micromolar concentrations of αS and compounds. The fact that a diverse group of polyphenols were analyzed enabled us to gain insight into particularly important structure–activity relationships. We also looked at the controversial issue of whether the anti-amyloid effects of polyphenols are related, or not, to their antioxidant or metal-ion chelation activities. The 15 phenolic compounds used in this study are listed together with their molecular weight (MW), calculated partition coefficient (c Log P) and sources. Refer to Fig. 4 for the chemical structures of the compounds. Expression and purification of recombinant αS was performed as described previously by Kostka et al. [25]. Briefly, pET-5a/α-Synuclein (136TAT) plasmid (wt-plasmid a kind gift by Philipp Kahle; the 136-TAC/TAT-mutation was performed by Matthias Habeck) was used to transform Escherichia coli BL21(DE3)pLys (Novagen, Madison, WI, USA), and expression was induced with isopropyl β-d-thiogalactopyranose (IPTG) for 4 h. Cells were harvested, resuspended in 20 mM Tris and 25 mM NaCl, pH 8.0 and lysed by boiling at 95 °C for 30 min in a water bath. The lysate was centrifuged at 17 000 g and 4 °C for 15 min and the supernatant was filtered and loaded into a HiTrap Q HP column (5 ml, GE Healthcare). After elution, it was ultra-centrifuged at 40 000 g and 4 °C for 45 min and the supernatant concentrated using Vivaspin 2 columns (MWCO 3 kDa). Afterwards the synuclein-fraction was gel filtrated over a Superdex 75 prep grade column (25 ml, GE Healthcare) with 20 mM Tris and 150 mM NaCl, pH 7.0 as running buffer to separate monomeric and oligomeric α-synuclein by size. The protein concentration was determined in a standard Bicinchoninic Acid (BCA)-solution assay. Aliquoted protein was stored at −80 °C after freezing in liquid nitrogen. Protein labeling by fluorophores was performed as previously described [25]. Briefly, αS was labeled with Alexa Fluor-488-O-succinimidylester (“green”) and Alexa Fluor-647-O-succinimidylester (“red”) (Invitrogen), respectively. Unbound fluorophores were separated by two filtration steps in PD10 columns (Sephadex G25; Amersham Biosciences). Quality control of labeled αS was performed by FCS measurements. To avoid repeated freeze/thawing, the purified recombinant fluorescently labeled monomeric αS stocks were divided into smaller aliquots of 5–10 μl, and stored at −80 °C. FIDA and SIFT measurements were carried out on an Insight Reader (Evotec-Technologies, Hamburg, Germany) with dual color excitation at 488 and 633 nm. Excitation power was 200 W at 488 nm and 300 μW at 633 nm. All measurements were performed at room temperature. The fluorescence signal was analyzed by FIDA using FCSPP evaluation software version 2.0 (Evotec-Technologies). For FIDA and SIFT analysis, fluorescence from the two different fluorophores was recorded simultaneously with two single-photon detectors; photons were summed over time intervals of constant length (bins) using a bin length of 40 μs. The frequency of specific combinations of “green” and “red” photon counts was recorded in two-dimensional intensity distribution histograms. Evaluation of SIFT data in two-dimensional intensity distribution histograms was performed by summing up the numbers of high intensity bins each of 18 equally sized sectors using a 2D-SIFT software module (Evotec-Technologies). For threshold setting, non-aggregated reference samples were used. This single molecule detection technology allows highly sensitive analysis of protein aggregation by changes in the brightness of individual particles. FIDA is able to distinguish between differently bright species and, as such, gives indirect information about particle sizes. A fivefold stock solution of fluorescently labeled αS was prepared by mixing αS labeled with Alexa-488 and αS labeled with Alexa-647. The concentrations of αS-Alexa-488 and αS-Alexa-647 were adjusted to approximately 10 molecules per focal volume and 15 molecules per focal volume, respectively. Aggregation was started by diluting the stock solution in 50 mM Tris–HCl buffer (pH 7.0) containing final 1% DMSO and 20–30 nM labeled αS in a total volume of 20 μl. FeCl3 was used at a final concentration of 10 μM. The aggregation reaction was typically complete in 10 min, but the plate was left for another 20 min at room temperature to minimize variance in measurements. All experiments were performed in 96-well-plates with a cover slide bottom. All polyphenol compounds and black tea extract (Table 1) were obtained from Sigma–Aldrich, Germany; except for scutellarein (Scut) which was purchased from Pharmasciences Laboratories, France. N-Acetyl-l-cysteine (N-acetylC), Vitamin C (VitC) and Desferrioxamine (Desf) were also obtained from Sigma. Compounds were dissolved in dimethylsulfoxide (DMSO) and kept as 10 mM stock solutions at −20 °C to maintain maximal stability. During the experiments they were used immediately after thawing and kept away from light. Compound screening was typically done at 10 μM (3 μg/ml for BTE) in a total assay volume of 20 μl. We initially confirmed that all compounds did not autofluoresce or quench at the diluted concentration (data not shown). To assay for inhibitory activity, compounds were added to monomeric αS in 50 mM Tris–HCl buffer (pH 7.0), before starting aggregation with 1% DMSO ± 10 μM FeCl3. To test for disaggregating activity, the compounds were added to pre-formed oligomeric αS. In both cases, after initiating aggregation the plate was left for ∼30 min at room temperature. The average SIFT signal from samples (in triplicate) was expressed as a percentage of the average signal from αS aggregation control wells. SIFT signals were generated by summing up the photon counts from the “green” and “red” channels for all bins above cut-off (threshold) level. Results were expressed as the means and the standard deviation (S.D.) values, with n as the number of experiments. Calculation of IC50 values from dose-dependency curves was performed using an on-line software facility (http://bsmdb.tmd.ac.jp:3000/cbdb/ic50). Differences between means were determined by unpaired Student's t test. In all analyses, the null hypothesis was rejected at the 0.05 level. Using three independent single particle-based methods, we recently described a robust in vitro multistep aggregation model for αS [25, 26]. In this model, incubation of nanomolar quantities of αS with 1% DMSO for 10 min was sufficient for maximal conversion of αS monomers into small aggregates, termed ‘type-I’ oligomers (Fig. 1 , upper panels). Cross-correlation analysis indicated a size of ∼20 monomers for these oligomers [25]. Addition of 10 μM Fe3+ ions in the aggregation assay resulted in the formation of larger (>100meres) ‘type-II’ oligomers (Fig. 1, upper panels). AFM measurements agreed with the confocal single particle fluorescence results [25]. Both type-I and type-II oligomer species are on-pathway to amyloid fibrils, as indicated by Thioflavin T assays. The robust αS aggregation model thus characterized is ideally suited for application in screening of small-molecule compound libraries for aggregation inhibitors. We first tested the ability of 13 polyphenols and black tea extract (Table 1) to antagonize the aggregation of alpha-synuclein into type-I oligomers (Fig. 2 A). The degree of inhibition by the various compounds was wide-ranging, from very strong inhibition (e.g. BTE = 0.4% and Mor = 0.7%), to strong (e.g. Myr = 10%, Quer = 12%), moderate (e.g. RA = 45%, Resv = 47%), and relatively weak (e.g. Gen = 68.6%, Api = 65.1%); this feature would later prove useful in deducing structure–activity relationships. Nine compounds – namely, BTE (see also Fig. 1, lower panels), Mor, TA, Baic, NDGA, EGCG, Gink, Myr and Quer – blocked type-I αS oligomer formation to less than 20% of the control value, the latter representing aggregation in the presence of DMSO only. Dose-dependency curves were also performed on seven of these anti-aggregation compounds, using compound concentrations between 12.5 nM and 10 μM. The following IC50 values were obtained: TA, 61 nM; NDGA, 77 nM; EGCG, 0.79 μM; BTE, 0.81 μM; Baic, 2.03 μM; Myr, 3.57 μM; Mor, 4.24 μM (Fig. 7 ). Aggregation of αS occurs more strongly and results in larger type-II oligomers when ferric iron is added to the reaction mixture (Fig. 1). Thus, we selected the best ten compounds from the previous assay, and assessed their ability to antagonize αS aggregation in the presence of Fe3+ ions as well. A broad range of inhibition was again observed, with six compounds blocking type-II αS oligomer formation to lower than 20% of the control value (Fig. 2B): Baic (2.62 ± 0.6%), NDGA (7.96 ± 1.1%), Myr (10.12 ± 0.7%), EGCG (10.13 ± 1%), BTE (15.9 ± 4.4%), and TA (18.9 ± 1%). Hence, five polyphenols – namely, Baic, EGCG, Myr, NDGA and TA – as well as BTE could be selected as compounds that potently interfere with the assembly of alpha-synuclein into multimeric oligomers, with IC50 in the low micromolar range. Other studies have shown that polyphenols can induce oligomer formation. EGCG, for instance, promoted the folding of αS monomers into highly stable oligomers that were non-cytotoxic and off-pathway to fibrillogenesis [16]. Similarly, Baic induced the formation of, and stabilized, αS oligomers consequently preventing their fibrillization [21]. Nevertheless, very strong inhibition of αS oligomerization by the potent compounds was evident in our single-molecule approach. The 13 polyphenols and BTE listed in Table 1 (excluding scutellarein) were next tested for their ability to disaggregate pre-formed αS oligomeric structures, the latter induced either by DMSO alone (type-I; Fig. 3 A), or with the inclusion of ferric iron (type-II; Fig. 3B). It was found that Baic, EGCG, Myr, NDGA and BTE significantly disrupted both types of oligomers to <20% of their initial aggregated state. We also carried out dose-dependency experiments (0.01 – 10 μM) to determine IC50 values for disaggregation potency: Baic, 0.85 μM; EGCG, 0.26 μM; Myr, 2.52 μM; NDGA, 20 nM; TA, 67 nM; BTE, 0.79 μM (Fig. 7). Interestingly, the two compounds having the lowest IC50 values for disaggregation, NDGA (IC50 = 20 nM) and EGCG (IC50 = 260 nM), show “linear” dose–response curves. This contrasts with the sigmoidal shapes of the other compound curves. An effect was observed even at the lowest concentrations of 10 nM for NDGA and 100 nM for EGCG, with a linear increase (on a logarithmic scale) to almost 100% effect at 10 μM. The implication is that disassembly of the αS aggregate structures occurred even at concentrations just above zero and increased with the concentration of the compound. At this stage, we therefore concluded that Baic, EGCG, Myr, NDGA and BTE could be classified as being the best combined inhibitors and disaggregators of αS oligomers. On the other hand, compounds like purpurogallin trimethyl ether (Purp), Api, Gen and Resv, which were relatively weak inhibitors, were also poor disaggregators of DMSO-induced synuclein oligomers. With regards to Fe-induced oligomers, Gink and RA were similarly weak in both inhibition and destabilization of larger type-II synuclein species. Thus, a pattern was emerging suggesting an important structure–function relationship: the ranking of overall compound potency (i.e. inhibition and disaggregation) in our aggregation model could be fairly accurately predicted based upon the number of hydroxyl groups present on a single phenyl ring, as follows: trihydroxy- > dihydroxy- > monohydroxy-phenyl ring (refer to Fig. 4 for grouping of compounds according to their chemical structures). The favouring of trihydroxyphenyl rings is made especially clear when comparing Baic and Api, the latter being much weaker in inhibiting or disaggregating type-I synuclein oligomers. Structurally, both have the same flavone structure and both possess a total of three –OH groups. The difference lies in that whilst Baic has all three –OH groups attached to the same benzene ring, Api has a dihydroxyphenyl at one end of the molecule and a hydroxyphenyl ring at the other end (Fig. 4). Therefore, since both compounds have a total of three hydroxyl groups it can be inferred that it is not the total number of –OH groups present in the molecule that is important for anti-aggregate activity, but rather the distribution of the –OH groups, most significantly the presence of three vicinal hydroxyl groups as in baicalein. To test this hypothesis another polyphenol with a trihydroxyphenyl ring, Scutellarein (Scut; Fig. 4), was selected for direct comparison to Baic and Api. Indeed, in agreement with the structure–activity hypothesis outlined above, Scut behaved in a similar way to Baic with regards to both inhibition of type-II synuclein oligomers and disaggregation of pre-formed oligomers, whereas Api was much less effective (Fig. 5 ). One can point out the apparent exception of NDGA to this insight into structure–activity relationship. NDGA, which does not have three adjacent –OH groups, is nonetheless one of the most effective compounds overall. However, NDGA has a very symmetrical structure with dihydroxyphenyl rings at both ends of the molecule; perhaps the structural symmetry confers an advantage with regards to possible orientations when binding to the αS molecules. Our data is in agreement with a related finding that dopamine and derivative compounds that have vicinal dihydroxy groups were shown to be effective inhibitors of αS fibrillization [9]. An issue we wanted to address was whether the well-known antioxidant, and even metal-ion chelating, activities of polyphenolic compounds could be to their observed anti-aggregation We found inhibitory by the μM) and or by the powerful iron μM) on αS aggregation induced by DMSO and Fe3+ ions (Fig. ). Similarly, disassembly of type-II oligomers was not evident by 100 μM ± or ± to the In it appears that antioxidant or are for the effects of polyphenols used at 10 μM) on αS is between our compound of anti-amyloid effects and data on the of the Using confocal single-molecule fluorescence we have the effects of phenolic compounds and black tea extract on αS oligomerization induced by We that a group of small-molecule polyphenols having three vicinal hydroxyl groups Myr, together with NDGA and black tea a of compounds that can potently αS aggregation into oligomers as well as disaggregate pre-formed αS oligomers, with IC50 in the μM) range. The concentration of αS is to be μM in human and in Similarly, it is not that polyphenols or their can in the than 10 μM For these we that the from our study, as they are to the in vivo by structure–activity analysis, we that the of potent compounds that would binding to the of the αS and hydroxyl groups the presence of three > two > one –OH groups on the same ring that would the of the and/or For instance, molecular recently showed that Myr amyloid oligomers by it be that the polyphenols can also to other of the αS For et al. have that the polyphenol a of dopamine fibrillation of αS by binding to the the of polyphenolic compounds in of it has to be first out that polyphenols increased and lower the The of a high concentration in repeated of the polyphenols over whilst the activities of the from the compounds in a study in black tea showed an with Parkinson's disease that was not by other like total intake or of the phenolic compounds myricetin and NDGA, both of which were found to have potent in our was in amyloid aggregation in vivo and the development of in potent polyphenol out in our study, EGCG, has been in water to transgenic in a and Hence, it be a that with polyphenols an effective treatment for neurodegenerative disorders of the amyloid of EGCG in can be increased more than by formation and of EGCG EGCG, and other polyphenols, to effective concentrations in the from our study, together with from previous a strong for the of a group of polyphenols in transgenic of Thus, EGCG, NDGA and black tea extract are all for with to inhibition of αS oligomer formation in the latter representing the most molecular species of αS into the structure–activity of natural outlined above the of in Parkinson's disease which possess in vivo (e.g. ability to the but which the of the natural This was by from the the and and the