Abstract
The misfolding of amyloid-beta (Aβ) peptide is one of the pathological hallmarks of Alzheimer's disease (AD). Polyphenols are strong antioxidants and metal chelators, with characteristics that are of beneficial therapeutic values for their development as candidates targeting neurodegenerative and metal-induced diseases. We have demonstrated here the electrochemical properties of a green tea component, (−)-epigallocatechin-3-gallate (EGCG), and its potent activity on Aβ peptides. Characterization of early interactions (≤48 h) between EGCG and Aβ was conducted using square wave voltammetry (SWV). The interaction of Cu(ii) ions with the Tyr-10 residue of Aβ was shown to be affected by surrounding His residues. Morphological changes due to the binding of EGCG and Cu(ii) were also elucidated using transmission electron microscopy (TEM). Electroanalytical techniques are promising for facilitating the investigation of metals and flavonoids in drug screening studies.
Electrochemical properties of a green tea component (−)-epigallocatechin-3-gallate (EGCG) have been studied to demonstrate its potent antioxidant and metal-chelating activities on Aβ peptides.
Introduction
Alzheimer's disease (AD) is a pathological condition originating from a slow progression of irreversible neuronal cell loss within various regions of the brain.1 It is at present without a cure or any effective early diagnostic means. Available treatments are all palliative and provide little symptomatic relief with severe side effects.2 To date, AD alone affects approximately 2% of the world population.3,4 13% of people in the US over the age of 65 are afflicted, with the figure rising rapidly to over 40% for those over 85 years of age.5 These numbers are expected to rise even faster with longer life expectancy of the current generation.5
Amyloid-beta (Aβ) has been reported to be the major constituent of senile plaques implicated in pathological deterioration in AD.6,7 The fibrillar structure of Aβ aggregates has been investigated using electron microscopy, X-ray diffraction, and light scattering photometry, revealing the two most abundant isoforms as Aβ(1–40) and Aβ(1–42).6 Due to its higher hydrophobicity, Aβ(1–42) is reported to aggregate at a more rapid rate than Aβ(1–40).2
Metals play crucial roles in neural transmission, oxygen transport and synthesis/metabolism of neurotransmitters in the brain.7,8 On the other hand, the dysregulation of metals can also result in oxidative stress and disorders. For instance, copper has been studied in an attempt to understand its adverse effects in numerous metal-mediated diseases.9,10 Oxidative stress is often associated with cell/tissue damage leading to heart attacks, stroke and cancers.11,12 For AD patients, a 3–5 fold increase in the levels of iron and copper have previously been reported.13 Toxic deposits were detected after incubation of Aβ oligomers with Cu(ii) ions in vitro.14,15 In view of these findings, there is an urgent need for the development of new and cost-effective techniques to explore novel therapeutic agents accelerating the treatment of metal-related diseases.
Flavonoids are naturally occurring molecules that are found in large quantities in plants.16,17 With increasing consumption of flavonoid containing foods and the growing popularity of green tea all over the world, the health benefits of flavonoids have captured the attention of researchers in diverse fields.18–20 Recent studies have indicated that flavonoids have broad-spectrum pharmacological activities and extensive biological effects.18–20 Currently, more than 8000 flavonoids have been identified, with a majority of them belonging to a group of polyphenols. Their structure is often characterized by a ring system with numerous hydroxyl groups, which give flavonoids their antioxidant and metal chelating abilities (Fig. 1).18,19 Recently, it has been shown that the flavonoid, (−)-epigallocatechin-3-gallate (EGCG), can directly remodel large and matured fibrils of Aβ into smaller and amorphous aggregates, which are non-toxic to cells.20–25 It was hypothesized that EGCG could potentially slow down the progression of AD by providing relief to oxidative stress or indirect chelation of metal ions for potential aggregate removal.20–25
Structure of EGCG with the possible metal coordination sites circled.
In this report, the interaction of EGCG with Aβ(1–40) and Cu(ii) was examined using square wave voltammetry (SWV). Detection was based on the oxidation of the Tyr-10 residue in Aβ(1–40).26–31 As the peptide aggregated, the extent to which the Tyr residue was exposed to the electrode surface decreased, changing the oxidation current signal. This method was extremely useful in monitoring not only the aggregation state of Aβ(1–40), but also the progression of its structural and conformational changes together with other species at the early stages of aggregation. Comparative studies on flavonoid–metal interactions between EGCG and Cu(ii) were performed under various concentrations, incubation periods and solution conditions. UV-vis spectrophotometry was also employed to study the chelation of Cu(ii) with EGCG.32–34 The findings obtained by the electrochemical and optical studies in this report can accelerate drug discovery efforts, with the well-established anti-aggregation and metal chelating antioxidant properties of EGCG being readily applied in future rational drug design studies towards AD therapy.
Experimental
Chemical reagents
Cu(ii) chloride (ACS reagent, ≥99.0%) and (−)-epigallocatechin-3-gallate (EGCG, from green tea, ≥95.0%) were purchased from Sigma-Aldrich (St. Louis, MO). Ultrapure water generated by the Cascada LS (Pall Co., NY) water purification system at 18.2 MΩ was used in the preparation of all the solutions. Other reagents were of analytical grade and used as received.
Apparatus
Electrochemical measurements were performed using a μAutolabIII potentiostat and operated using the General Purpose Electrochemistry Software (GPES) (Metrohm, Switzerland). Measurements were conducted using carbon screen-printed electrodes (geometric area: 2.64 mm2, carbon counter-electrode, and Ag/AgCl reference electrode) from Biodevice Technology Ltd. (Kanazawa, Japan). A Nanodrop 2000c spectrophotometer (Wilmington, DE) with cuvette capability was employed to perform UV-vis spectroscopy. Temperature-dependent incubations were performed using a VWR Symphony incubator (Grafton, MA).
Procedure
Square-wave voltammetry
Stock solutions of EGCG were prepared fresh in ultrapure water at a concentration of 1 mM. Aβ(1–40) stock solutions were prepared in a mixture of dimethyl sulfoxide (DMSO) and 50 mM phosphate buffer saline solution (PBS) at pH 7.4 with a concentration of 100 μM.27,35 Samples were prepared by incubating 100 μM concentration of each species with required dilutions using PBS in the following combinations: (1) Aβ(1–40) with EGCG, (2) Aβ(1–40) with Cu(ii), and (3) Aβ(1–40) with EGCG and Cu(ii). In addition, blank samples were prepared from stock solutions of each reagent. All the samples were protected from light, and incubated at 37 °C with constant shaking. For each measurement, three aliquots of 10 μL samples were removed from the incubated solutions, and transferred to vials containing 90 μL of PBS to achieve the desired concentration of 10 μM. The samples were applied onto the surface of the carbon screen-printed electrodes in a 20 μL drop. Prior to any analysis, sample vials were mixed to ensure homogeneity. Electrochemical oxidation signals were measured using SWV (applied potential from 0.1 to 1 V vs. Ag/AgCl; step potential, 5 mV; amplitude, 25 mV). Pencil graphite electrodes (Pentel, Japan) were used for the EGCG and Cu(ii) interaction studies. The pencil graphite was used as the working electrode together with a Ag/AgCl reference electrode and a Pt counter electrode. Similarly, SWV was performed with a 5 mV step potential at 10 mV amplitude, scanning from −0.35 to 0.80 V.
The raw SWV voltammograms were treated with the Savitzky–Golay smoothing (level-4) feature of GPES. The baseline correction feature was set to a peak width of 3 mV. After each measurement, the used working electrodes were disposed, thus, non-specific adsorption and cross-contamination between electrodes could be suppressed. Error bars indicate the standard deviation of signals obtained from triplicate measurements on separate electrodes (n = 3).
UV-vis spectrophotometry
Stock solutions of flavonoids (1 mM) and metals (5 mM) were diluted in various buffer solutions. The samples were transferred to 3 mL cuvettes, and measured using the Nanodrop 2000c spectrophotometer. Interactions of flavonoids and metals were observed by following changes in the characteristic wavelengths, and plotted against the absorbance values. All measurements, except for temperature-dependent studies, were taken under ambient conditions (24 ± 1 °C). At the end of each measurement, cuvettes were rinsed thoroughly with ethanol, and twice with ultrapure water to avoid cross-contamination. Each measurement was repeated at least three times (n = 3).
TEM imaging
TEM images were prepared by absorbing 6 μL of each sample onto a nickel Formvar mesh grid (Electron Microscopy Sciences, Hatfield, PA) for 1 min. The grids were then negatively stained with 1% uranyl acetate for 1 min, then air dried. The resulting samples were viewed on a Hitachi H-7500 transmission electron microscope, which was operated at a range between 2 and 200 kV depending on the magnification required.
Results and discussion
The interactions of Aβ(1–40) with Cu(ii) ions were observed by analyzing changes in their electrochemical properties over time. Within the first 48 h, fluctuations were observed as a result of the exposure of the Tyr-10 residue of the Aβ(1–40) to the electrode surface during the early stages of aggregation. This was expected as it is a characteristic of Aβ(1–40) displaying a lag phase of up to 50 h, consistent with a previous report by Skaat et al.36 Electrochemical measurements of EGCG using SWV have previously shown two oxidation peaks between 0.40 and 0.50 V.37 When only EGCG was added to Aβ(1–40), a wide signal with two shoulder peaks was observed initially ranging from 0.35 V to 0.60 V (Fig. 2). The peak was remarkably high in comparison to that of the initial signal of Aβ(1–40) alone (0.40 V to 0.50 V). This increase was attributed to the oxidation of EGCG itself along with an overlapping peak of the Tyr oxidation from Aβ(1–40) (∼0.60 V).36 As time progressed, the signal from 0.35 to 0.40 V diminished dramatically. After 6 h of incubation, the electrochemical signal at ∼0.60 V diminished to an even smaller magnitude, indicating the interaction of EGCG with Aβ(1–40). The decreasing signal of EGCG was attributed to the intercalation of the planar aromatic structure of EGCG between the β-sheets formed by Aβ(1–40) oligomers at the early stages of aggregation. This possible interaction would result in the limited exposure of EGCG to the electrode surface, and thus, would generate low current signals. The decline of the Aβ(1–40) peak was attributed to the blocking of Tyr-10 during interactions with EGCG. A control experiment was performed to show that the current decrease was not caused by degraded EGCG residues, as no significant change in the peak potential and current of EGCG was observed during its incubation in PBS for 24 h (Fig. S1, ESI†). After 24 h, the current of the oxidation signal increased, suggesting that prolonged incubation triggered a conformational change that exposed the Tyr-10 residue to the electrode surface. We postulated that the formation of oligomers and protofibrils might have contributed to the shift in the signal from ∼0.50 to 0.70 V.
Square-wave voltammograms of 100 μM EGCG incubated with the same concentration of Aβ(1–40) at 37 °C in 50 mM PBS (pH 7.40) using a screen-printed carbon electrode.
The interaction of EGCG with Cu(ii) was also studied using SWV. A well-defined Cu(i) oxidation peak27 at ∼−0.10 V was found to decrease upon titration with EGCG. This was attributed to the chelation of the metal ion, reducing its availability for oxidation on the electrode surface (Fig. S2, ESI†). This control experiment confirmed the strong interaction between EGCG and Cu(ii). When Cu(ii) ions interacted with Aβ(1–40), a decrease in the Tyr oxidation current was observed for the first 6 h as shown in Fig. 3. This observation was attributed to the additional interaction of Cu(ii) ions with His residues (positions 6, 13 and 14) as previously reported.38 Possible interaction of one or more Cu(ii) ions with His residues might have resulted in a significant change in the conformation of the peptide, in a way that would have exposed Tyr-10 to the electrode surface (Fig. S3, ESI†).
Plot for the electrochemical Tyr oxidation signals of Aβ(1–40) blank, Aβ(1–40) with Cu(ii), Aβ(1–40) with EGCG and Aβ(1–40) with both EGCG and Cu(ii). Error bars were created based on the standard deviation for each set of triplicate measurements (n = 3).
Aβ(1–40) aggregation was also characterized in the presence of both Cu(ii) and EGCG (Fig. 3), and the results were in agreement with previous ThT studies.23,25 There was a significant decrease in the electrochemical signal, which indicated the chelation of Cu(ii) ions by EGCG in the first 2 h. The overall trend of the signal was comparable to that of Aβ(1–40) with EGCG, signifying the predominant effect of EGCG on aggregation over Cu(ii) ions.
In UV-vis studies, Cu(ii) induced a spectral change for all conditions when interacted with EGCG (Fig. 4A and B). This was in agreement with previous studies.39,40 When Cu(ii) interacted with EGCG at neutral pH, the characteristic EGCG peak at ∼270 nm started to diminish with increasing concentrations of Cu(ii). At the same time, a new peak appeared at ∼320 nm, and increased with the addition of Cu(ii) ions. This suggested that the EGCG–Cu(ii) complex was formed immediately after metal exposure. Fig. 4C and D shows the data obtained with temperature and time dependence studies, respectively. At 0 h, 250 μM Cu(ii) was added to 25 μM EGCG, and generated the same peak at ∼320 nm as shown in Fig. 4A. However, as time progressed, the peak at ∼320 nm gradually disappeared after 24 h. The effect of temperature was also clearly demonstrated. Instead of the total disappearance of ∼320 nm peak at 37 °C, absorbance peaks were still observed at room temperature after 24 h. This result was attributed to the effect of high temperature on the chelating ability of EGCG.
UV-vis absorption spectrum of (A) 25 μM EGCG with various concentrations of Cu(ii), (B) EGCG alone, (C) the incubation of 25 μM EGCG with Cu(ii) at 37 °C in 50 mM PBS, and (D) the incubation of EGCG with Cu(ii) in 50 mM PBS at room temperature.
Conformational changes of Aβ(1–40) upon interaction with EGCG and Cu(ii) were imaged using transmission electron microscopy (TEM) with samples incubated for 72 h. TEM images were in agreement with our electrochemical and optical results. Fig. 5A showed distinct fibrillar networks of the defined elongated structure for Aβ(1–40) alone. In contrast, the EGCG incubated sample was entirely devoid of these defined fibrillar structures and exhibited only globular aggregates with non-distinctive edges (Fig. 5C). We also noted the different density of the overlapping globular aggregates. The high density was proposed to be caused by the intercalation of EGCG to the β-sheets at the early stages. With only Cu(ii) present, the Aβ(1–40) aggregates showed a dense structure (Fig. 5B) due to possible interactions within the metal binding region of Aβ(1–40)peptides.38,41 Similarly, unstructured aggregates were observed, when both EGCG and Cu(ii) ions were incubated with Aβ(1–40) (Fig. 5D). This result suggested that the chelating ability of EGCG had impeded the formation of the Cu(ii)–His complex, and, thus, resulted in reduced fibril formation, in support of our electrochemical and optical observations. It was noted that these amorphous globular structures lacked defined fibrillar components, which were closely associated with the neurotoxicity of Aβ(1–40) fibrils. Such structures of EGCG and Aβ(1–40) were reported to be non-toxic, as their conformation deviated Aβ from the pathological pathways in AD.42–44
TEM images of Aβ(1–40) aggregation in vitro measured using (A) 100 μM Aβ(1–40) blank, (B) Aβ(1–40) incubated in the presence of 100 μM Cu(ii), (C) Aβ(1–40) incubated in the presence of 100 μM EGCG, and (D) Aβ(1–40) incubated with both EGCG and Cu(ii) at 100 μM under the same conditions at 37 °C in PBS for 4 days as described in the Experimental section.
Conclusion
The early interactions between EGCG, Aβ(1–40) and Cu(ii) ions were characterized using SWV. Electrochemical and optical data were further complemented by TEM images on Aβ oligomers. The results indicated the intricate nature of the interaction between EGCG and Aβ(1–40), and the possible binding of Cu(ii) with the Tyr and His residues of Aβ(1–40). It is important to understand the behaviour of flavonoids with Aβ(1–40) in an effort to discover their full potential as therapeutics. Moreover, we demonstrated that electrochemical analysis of amyloid fibril formation using SWV can provide an efficient platform with significant sensitivity for the study of early stages of protein aggregation in vitro. The reported results can be readily applied to various metal-related protein-misfolding diseases. It is envisaged that in vitro electroanalysis of peptide and metal interactions described in this report can empower future drug discovery efforts.
Acknowledgements
The authors gratefully acknowledge the financial support from the Alzheimer Society of Canada and NSERC Discovery Grant.
Notes and references
Footnotes
Electronic supplementary information (ESI) available: Additional experimental data and figures. See DOI: 10.1039/c3mt20106f
