Biochimica et Biophysica Acta 1865 (2017) 589–603

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbapap

Polyphenols in combination with β-cyclodextrin can inhibit and

disaggregate α-synuclein amyloids under cell mimicking conditions: A
promising therapeutic alternative

Saurabh Gautam a, Sandip Karmakar a, Radhika Batra a, Pankaj Sharma b, Prashant Pradhan b, Jasdeep Singh b,
Bishwajit Kundu b,⁎, Pramit K. Chowdhury a,⁎
a
Department of Chemistry, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India
b
Kusuma School of Biological Sciences, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India

a r t i c l e i n f o a b s t r a c t

Article history: Parkinson's disease is characterized by the presence of insoluble and neurotoxic aggregates (amyloid fibrils) of an
Received 25 July 2016 intrinsically disordered protein α-synuclein. In this study we have examined the effects of four naturally occur-
Received in revised form 17 February 2017 ring polyphenols in combination with β-cyclodextrin (β-CD) on the aggregation of α-synuclein in the presence
Accepted 22 February 2017
of macromolecular crowding agents. Our results reveal that even at sub-stoichiometric concentrations of the in-
Available online 24 February 2017
dividual components, the polyphenol–β-CD combination(s) not only inhibited the aggregation of the proteins
Keywords:
but was also effective in disaggregating preformed fibrils. Curcumin was found to be the most efficient, followed
α-Synuclein by baicalein with (−)-epigallocatechin gallate and resveratrol coming in next, the latter two exhibiting very sim-
β-cyclodextrin ilar effects. Our results suggest that the efficiency of curcumin results from a balanced composition of the phenolic
Polyphenols \\OH groups, benzene rings and flexibility. The latter ensures proper positioning of the functional groups to max-
Amyloid inhibition and disaggregation imize the underlying interactions with both the monomeric form of α-synuclein and its aggregates. The unique-
Parkinson's disease ness of β-CD was reinforced by the observation that none of the other cyclodextrin variants [α-CD and HP-β-CD]
used was as effective, in spite of these possessing better water solubility. Moreover, the fact that the combinations
remained effective under conditions of macromolecular crowding suggests that these have the potential to be de-
veloped into viable drug compositions in the near future. MTT assays on cell viability independently confirmed
this hypothesis wherein these combinations (and the polyphenols alone too) appreciably impeded the toxicity
of the prefibrillar α-synuclein aggregates on the mouse neuroblastoma cell lines (N2a cells).
© 2017 Elsevier B.V. All rights reserved.

1. Introduction characterized by the presence of macroscopic amyloid deposits of the

A number of human pathological diseases such as Alzheimer's dis-
ease, type 2 diabetes, Parkinson's disease, Amyotrophic lateral sclerosis
(ALS), the prion diseases, and Huntington disease arise from the un-
wanted misfolding, oligomerization and aggregation of proteins [1].
Most of these diseases are characterized by the deposition of amyloid fi-
brils in body tissues. Among them Parkinson's disease is one of the most
common diseases with 7–10 million people worldwide suffering from
it. Parkinson's disease and dementia (due to synucleinopathies) are
Abbreviations: β-CD, β-cyclodextrin; CR, congo red; EGCG, (−)-epigallocatechin
gallate; HP-β-CD, (2-Hydroxypropyl)-β-cyclodextrin; IPTG, isopropyl β-D-
thiogalactopyranoside; SEC, size exclusion chromatography; TEM, transmission electron
microscopy; ThT, Thioflavin T.
⁎ Corresponding authors.
E-mail addresses: bkundu@bioschool.iitd.ac.in (B. Kundu),
pramitc@chemistry.iitd.ac.in (P.K. Chowdhury).
protein α-synuclein known as Lewy neuritis and Lewy bodies in brain
tissues [2].
α-Synuclein is copiously present in human brain tissues and also in
some other tissues such as red blood cells [2]. It is an intrinsically disor-
dered protein, composed of 140 amino acid residues that can be struc-
turally divided into three distinct segments (Fig. S1, Supporting
Information) namely: (a) N-terminal amphipathic segment (1–60
amino acid residues), (b) a hydrophobic central region (61–95 residues)
and (c) C-terminal acidic region (96–140 residues) [3]. The N-terminal
region shares homology with lipoproteins (with amphipathic α-heli-
ces) and contains a characteristic consensus hexameric sequence
(KTKEGV) which is repeated about four times. The central hydrophobic
region is known to be responsible for the aggregation of α-synuclein,
without which the protein loses its propensity to form amyloids. The
carboxy terminus region is highly acidic in nature and has been found
to interfere in the formation of aggregates due to its role as an

http://dx.doi.org/10.1016/j.bbapap.2017.02.014
1570-9639/© 2017 Elsevier B.V. All rights reserved.
590 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
intramolecular chaperone [4]. Three tyrosine residues, which are highly
conserved, are also present in this C-terminal region. In vivo, higher ex-
pression level of α-synuclein, increases its tendency to aggregate [5].
The actual function of this protein is ambiguous with its role in various
cellular activities having been reported, that includes interaction with
cell membranes/lipids, release of neurotransmitters and maintenance,
storage and transport of vesicles across cells [2,6]. In presence of alco-
hols, α-synuclein is known to form ordered structures such as α-helix
and β-sheet with low concentrations of ethyl alcohol enhancing its ag-
gregation [7]. Soluble oligomers apart from fibrils of α-synuclein have
also been implicated in neurotoxicity during Parkinson's disease [2].
Numerous small molecule inhibitors (e.g. flavonoids, dopamine, car-
bohydrates, antibiotics, synthetic chemical compounds, natural plant
products etc.) of amyloid formation by α-synuclein have been reported
in literature [8,9]. Polyphenols also constitute an important class of mol-
ecules that have been shown to inhibit and in some cases even disaggre-
gate amyloids formed by α-synuclein [10–20] and other proteins [21,
22]. Moreover, polyphenols are known to have anti-cancer, anti-oxida-
tive and anti-microbial properties thereby providing several additional
medicinal benefits. They are also reported to prevent cardiovascular dis-
eases and control cell death [23]. Recent studies on the mechanism of
inhibition and disaggregation of α-synuclein amyloids by polyphenolic
compounds have shown that the presence of two or more ‘vicinal
polyhydroxyphenyl groups’ is more important rather than the total
number of hydroxyl groups [10,13]. Molecular dynamics studies have
also pointed towards the role of these \\OH groups in inhibition and
disaggregation of amyloids [21,24]. However, curcumin, a naturally oc-
curring polyphenol derived from turmeric (Curcuma longa), is a unique
molecule with only monohydroxyphenyl rings but still being a highly
potent aggregation inhibitor [25–28]. It has been hypothesized that
curcumin binds to the aliphatic residues present in the middle hydro-
phobic stretch of native monomeric α-synuclein [29]. After this initial
interaction, the aromatic rings of curcumin bind to nearby hydrophobic
residues of α-synuclein. The same study proposed that curcumin
strongly prevents the formation of oligomers and fibrils of α-synuclein
by accelerating the intramolecular diffusion of the protein. This leads to
an increase in the reconfiguration rate of α-synuclein thereby limiting
the exposure time of the hydrophobic patches. Another study has
shown that curcumin reduces the toxicity of α-synuclein by binding
to it and minimizing the availability of hydrophobic surface [11].
A recent study from our group has demonstrated that curcumin in
combination with β-cyclodextrin (β-CD) not only inhibited but also dis-
aggregated α-synuclein amyloids in vitro [30]. We hypothesized that
curcumin, with its hydrophobic region and \\OH groups, interacts
with the hydrophobic middle region and hydrophilic residues of α-sy-
nuclein, respectively. β-CD then further helps by binding to aromatic
residues such as phenylalanine present in α-synuclein. β-CD is a cyclic
polysaccharide with seven glucose units. It is a bucket shaped molecule
with a hydrophobic internal cavity and hydrophilic exterior [31]. β-CD
itself is a biologically active molecule and has been used in various ap-
plications such as protein folding as a pseudochaperone [32], catalysis
[33], drug delivery [34], lowering of cholesterol [35] and inhibiting pro-
tein aggregation [36], to name a few. β-CD interacts with polyphenols
and helps in overcoming their limitations such as low bioavailability, in-
solubility, instability [37] and also increase their activity, efficiency [30],
anti-inflammatory and anti-proliferative properties [38].
Polyphenolic compounds are not always effective in vivo due to var-
ious reasons [39]. One of the major factors that add to the intracellular
complexity is the congested cellular interior. The intracellular environ-
ment is highly crowded with the concentration of macromolecules
ranging from 50 g/L to as high as 400 g/L [40]. Such an environment
has been shown to affect protein folding, structure, function, aggrega-
tion and other thermodynamic properties primarily via the excluded
volume effect [41]. Hence, the experiments carried out in dilute buffer
solutions may or may not always provide a reliable picture of the in
vivo environment. To address this issue, researchers have used both
synthetic and protein based crowders in an attempt to mimic the phys-
iological milieu [40,41]. In the current study, we have tried to extend the
use of the combination of polyphenols and β-CD to such cell mimicking
conditions. Based on our observations, we have also attempted to pro-
vide molecular insights into the underlying mechanism and role of
\\OH groups present in these polyphenols with regards to their ability
to modulate protein aggregation. We have used four different naturally
occurring polyphenols namely, resveratrol (from red grapes), baicalein
(from the Chinese herbal medicinal plant Scutellaria baicalensis) and
(−)-epigallocatechin gallate (EGCG) (from green tea) along with
curcumin which is obtained from turmeric (C. longa) (Fig. 1). The mac-
romolecular crowding agents used in this study are: Dextran 6, Dextran
40 and Ficoll 70 with an average molecular weight of 6 kDa, 40 kDa and
70 kDa, respectively. Our results reveal that these polyphenols in an
optimized combination with β-CD not only precluded formation of
aggregates but also disaggregated amyloids formed by α-synuclein
with varying efficiencies in the presence of the macromolecular
crowding agents. Moreover, the pronounced toxicities displayed by
the α-synuclein aggregates were significantly diminished when the
cells were treated with these polyphenol–β-CD combinations.
Taken together our data suggest that these synergistic combinations
may facilitate the long awaited expedition of such polyphenols to
drug molecules, though views against the same have been put for-
ward recently [42].
2. Materials and methods
2.1. Materials
Curcumin, resveratrol, baicalein, EGCG, thioflavin T (ThT) and congo
red (CR) were purchased from Sigma-Aldrich (St. Louis, MO, USA) and
used as received. Ethanol was procured from Merck (Mumbai, India).
Ampicillin, LB medium for bacterial growth and isopropyl β-D-thio-
galactopyranoside (IPTG) were purchased from Himedia chemicals
(Mumbai, India). All other chemicals used were of highest purity avail-
able/analytical grade.
2.2. Expression, isolation and purification of human α-synuclein
Human α-synuclein was expressed, isolated and purified with a sim-
ilar process as described earlier [30,43]. Plasmid pT7-7 containing the
gene for human α-synuclein was first transformed into E. coli BL21
(DE3) cells using calcium chloride. Then a single colony was picked
and inoculated into 100 mL LB medium enriched with 100 mg/mL am-
picillin. The culture was incubated at 37 °C and the absorbance at
600 nm was recorded at regular intervals. Induction was carried out
by adding IPTG (1 mM final concentrations) when absorbance reached
0.7 to 1.0. The cells were harvested and resuspended in 0.75 mL of buffer
(50 mM Tris-HCl, pH 7.5, 10 mM EDTA and 150 mM NaCl) and frozen at
−80 °C. Tubes with frozen cells were placed in a boiling water bath for
7 min and the supernatant was then collected after centrifugation at
maximum speed for 5 min. Streptomycin sulfate (136 μL/mL of 10% so-
lution per mL of supernatant) and glacial acetic acid (228 μL/mL of su-
pernatant) were added and centrifuged for 2 min. Again the
supernatant was recovered and precipitated with ammonium sulfate
(saturated ammonium sulfate at 4 °C was used 1:1, v/v, with superna-
tant). The protein was collected as a precipitate by centrifugation and
washed once with 1 mL of ammonium sulfate solution (4 °C, 1:1, v/v,
saturated ammonium sulfate and water). The washed pellet was resus-
pended in 900 μL of 100 mM ammonium acetate (to form a cloudy solu-
tion) and precipitated by adding an equal volume of ethanol at room
temperature. Precipitation with ethanol was repeated once more. The
pellet was resuspended in 100 mM ammonium acetate and extensively
dialyzed against 10 mM Tris-HCl buffer, pH 7.4.
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
Fig. 1. Chemical structures of the four different polyphenols [curcumin, resveratrol, baicalein and (−)-epigallocatechin gallate (EGCG)], β-cyclodextrin (β-CD) and two types of
macromolecular crowding agents (Dextran and Ficoll) used in this study. The structures were drawn using ChemDraw.
2.3. Preparation of solutions of polyphenols
Stock solutions of all four polyphenols (curcumin, resveratrol,
baicalein and EGCG) were prepared in ethanol. All the compounds
were added to protein solutions so that the final concentration of etha-
nol remained equal to or below 7.5% (v/v) which was required for pro-
tein aggregation.
2.4. Assays to monitor amyloid fibrils (fibril formation assay)
Amyloid fibrils formed by α-synuclein were monitored by using two
most common techniques namely: ThT based fluorimetric assay [44]
and congo red (CR) assay [45]. The protein solution with a concentra-
tion of 35 μM in 10 mM Tris-HCl buffer, pH 7.4 (containing 0.1 M
NaCl) was stirred with a micro-stirring bar in a glass vial at 37 °C with
or without different concentrations of macromolecular crowding agents
and ethanol, in the presence or absence of different concentrations of
polyphenols and/or β-CD. At regular intervals, aliquots of 60 μL were
taken out from the aggregating/disaggregating reaction mixtures
which contained 35 μM of protein in 10 mM Tris-HCl buffer, pH 7.4
apart from respective concentrations of ethanol and/or macromolecular
crowders. The aggregating/disaggregating mixtures also contained
polyphenol–β-CD combinations during the inhibition of aggregation
and disaggregation experiments. The aliquots of 60 μL from aggregat-
ing/disaggregating reaction mixtures were then mixed with 440 μL of
23 μM stock solution of ThT/CR (stock solution of ThT/CR was in
591
10 mM Tris-HCl buffer, pH 7.4) resulting in a total volume of 500 μL
with the final concentration of ThT/CR being 20 μM [The aliquots were
used directly without ThT/CR for recording circular dichroism spectra,
size exclusion chromatography (SEC) and transmission electron micros-
copy (TEM) experiments]. This resulted in a dilution of the aggregating/
disaggregating reaction mixture containing the protein and
polyphenol–β-CD combination. The dilution of the aggregating/disag-
gregating reaction mixture was 8.33 times (60 μL reaction mixture
+ 440 μL ThT in buffer = 500 μL / 60 μL = 8.33) thereby resulting in di-
lution of polyphenols as well by a factor of 8.33 present in the original
aggregating/disaggregating reaction mixture. ThT fluorescence acquisi-
tions were then performed by exciting the samples at 430 nm with an
emission at 485 nm using an Edinburgh FLS920 spectrofluorimeter (Ed-
inburgh Instruments Ltd., Livingston, UK). The fluorescence was mea-
sured using a quartz cuvette of 10 mm path length (Starna Scientific
Ltd., UK) with excitation and emission slit widths kept at 2 and 5 nm, re-
spectively. Relative intensity was obtained by calculating F/Fm, where F
is the fluorescence intensity of ThT in the presence of α-synuclein olig-
omers/aggregates and Fm is the fluorescence intensity of ThT in pres-
ence of monomeric native α-synuclein. All fluorescence
measurements were corrected for buffer contributions. CR absorbance
(at 540 nm) was recorded using a spectrophotometer (UV-2600,
Shimadzu, Japan) after incubation for 10 min at 25 °C. The absorbance
of protein only as a control in the absence of CR was also recorded by
taking an aliquot (60 μL) of protein sample with 440 μL of Tris-HCl buff-
er; similarly absorbance of CR only in the absence and presence of native
monomeric α-synuclein as a control was also recorded with 20 μM final
592 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
concentration of CR in Tris-HCl buffer. All the experiments were carried
out in triplicate and the error was b 5%. For disaggregation experiments,
preformed aggregates were obtained as described above and then the
compounds were added. Aliquots were then taken out at regular
intervals.
Aggregation and disaggregation kinetics of α-synuclein showed sig-
moidal behaviour reminiscent of nucleation dependent pathway and
were fitted in a sigmoidal curve using OriginPro 8.5:
A2−A1
Y ¼ A1 þ
1 þ 10ðLOGt 0 −t Þp
where Y is ThT fluorescence intensity or CR absorbance, A1 is Y value at
the bottom of the curve or bottom plateau, A2 is Y value at the top of the
curve or top plateau, LOGt0 is the time (in min) when the response is
halfway between Bottom and Top (half maximum) and p is the Hill
slope or Hill coefficient.
2.5. Size exclusion chromatography (SEC)
SEC experiments were carried out using an AKTA purifier FPLC sys-
tem (GE Healthcare). The column used was Superdex 200 Increase 10/
300 GL column (I.D.: 1.0 cm × 30 cm). The elution was carried out
using 10 mM sodium phosphate buffer (pH 7.4) with 0.1 M NaCl with
a flow rate of 0.50 mL/min. All the samples were injected for chromato-
graphic separation after centrifugation at 10,000 g for 20 min and filtra-
tion through a 0.22 μm membrane filter.
2.6. Circular dichroism spectroscopy
Far-UV circular dichroism spectra over the wavelength range of 195
to 250 nm were recorded on Jasco J-815 spectropolarimeter (Jasco Cor-
poration, Japan) equipped with a temperature controller. All the spectra
were accumulated at 25 °C using a cell with a path length of 0.1 cm and
corrected for buffer contributions with five scans averaged for each
spectrum. Circular dichroism data are presented in terms of normalized
ellipticity as a function of wavelength.
2.7. Transmission electron microscopy (TEM)
Aliquots (5 μL) were taken out from different α-synuclein samples
undergoing aggregation/inhibition of aggregation/disaggregation and
placed on a carbon-coated Formvar grid (Ted Pella, Inc., Redding, CA,
USA) and incubated for 5 min. Thereafter, the grids were washed with
MilliQ water and negatively stained with 2% (w/v) aqueous phospho-
tungstic acid (PTA) solution. PTA solution was freshly prepared in MilliQ
water and filtered through 0.22 μm sterile syringe filters before use.
Electron micrographs were recorded digitally on a FEI Morgagni 268D
instrument operated at 80 kV using magnifications in the range of
20,000–40,000X.
2.8. Cytotoxicity assay
The toxicity of various polyphenol-mediated α-synuclein species
were tested on N2a cells. Cells were cultured in DMEM supplemented
with 10% FBS and 1% penicillin-streptomycin (Cell clone, Genetix bio-
tech Asia Pvt. Ltd., India) and incubated at 37 °C and 5% CO2. The cells
were seeded in 96 well plates at a density of 104 cells and were allowed
to adhere and grow till the confluency reached to 85–90%. Experiments
were set in triplicates on two 96 well plates. The viability of cells was an-
alyzed using MTT assay that determines the mitochondrial dehydroge-
nase activity as described previously [46]. Only the viable cells reduce
MTT into formazan crystals by using the mitochondrial dehydrogenase
enzyme. To these different cells, components were added in the given
concentrations as indicated and incubated for 12 h. Cells incubated with
buffer alone having final ethanol concentration of b1% (~0.35%, v/v) and
with different concentrations of curcumin (7.5 μM), resveratrol (15 μM),
baicalein (10 μM) and EGCG (15 μM) served as positive controls. Cells in-
cubated with α-synuclein prefibrillar oligomers having initial concentra-
tion of 0.5 mg/mL (or 35 μM) for monomer was served as negative
control. Test samples of α-synuclein samples having concentration of
0.5 mg/mL (or 35 μM for monomer) drawn at the native (0 min), oligo-
meric (30 min), prefibrillar (50 min) and filamentous fibrillar (120 min)
states with or without optimized concentrations of polyphenols
(curcumin, resveratrol, baicalein and EGCG) and/or β-CD were added
and incubated for 12 h. The effective concentrations for curcumin, resver-
ð1Þ
atrol, baicalein and EGCG in combination with β-CD were 7.5 μM (with 15
μM β-CD), 15 μM (with 30 μM β-CD), 10 μM (with 20 μM β-CD) and 15
μM (with 30 μM β-CD), respectively. After incubation, MTT (sigma Al-
drich) was added at a concentration of 1 mg/mL and further incubated
at 37 °C and 5% CO2 for 2–4 h. DMSO was added to the wells and plates
were read at 570 nm (multiskan Go, Thermoscientific, USA). The resulting
absorbance values were converted to percentage viability with respect to
the untreated control cells. The statistical analysis for this data was done
by paired t-test method.
2.9. All atomistic molecular dynamics simulations
Molecular dynamics simulation was done using GROMACS 5.1.1 [47]
employing GROMOS 54a7 all-atom force field using periodic boundary
condition [48]. The starting model (PDB id: 2kkw, Model 1) was solvat-
ed in a periodic box with SPC water model. The initial structures of test
compounds were imported from PubChem Compound Database while
the topology and other parameters were assigned using ATB topology
builder [49,50]. In separate setups for α-synuclein, simulations were
carried out in the absence and presence of polyphenols. Briefly, the
monomers of the protein along with five monomers of individual poly-
phenols were placed in random orientation in individual cubical boxes.
Counter ions were added to the solvent to neutralize electrical net
charge of the system. The system was then minimized for 50,000
steps using a steepest decent algorithm followed by an equilibration
run of 100 ps in NVT (constant Number of particles, Volume, and Tem-
perature) ensemble with restraints on the protein atoms. The NPT (con-
stant Number of particles, Pressure, and Temperature) ensemble was
used for production simulation. Systems were simulated at 310 K, main-
tained by a Berendsen thermostat [51], with a time constant of 1 ps,
with the protein and non-protein molecules coupled separately. Pres-
sure coupling was done employing a parrinello-rahman barostat [52]
using a 1 bar reference pressure and a time constant of 2.0 ps with com-
pressibility of 4.5e−5 bar using isotropic scaling scheme. Electrostatic
interactions were calculated using the Particle Mesh Ewald (PME) sum-
mation. The production run was performed for 80 ns each with 2 fs time
step in every run. The coordinates were recorded at each 10 ps interval.
The resulting trajectories were analyzed by inbuilt GROMACS tools
while the images were created using PyMol [53] package. Secondary
structure analysis was done using VMD package [54].
3. Results
3.1. Aggregation
Aggregation of α-synuclein was carried out in presence of the mac-
romolecular crowding agents and was monitored using ThT (Thioflavin
T) based fluorescence [43] and CR (congo red) binding assays [44]. Both
these dyes are commonly used for following the progress of aggregation
and have been shown to complement other well established tech-
niques. Moreover due to concerns over ThT fluorescence data interpre-
tation in presence of the polyphenols [55], we have also used other
techniques to further validate our findings such as circular dichroism,
transmission electron microscopy, and size exclusion chromatography
apart from congo red absorbance.
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603 593

Fig. 2. Aggregation profiles of α-synuclein monitored using ThT fluorescence (λex =
430 nm and λem = 485 nm) in the presence of different concentrations of the
macromolecular crowding agents namely, Dextran 40, Dextran 6 and Ficoll 70.
Aggregation was carried out at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing
0.1 M NaCl with the protein concentration being kept at 35 μM. The data plotted are a
mean of the experiments done in triplicate and the standard deviation so obtained has
been included for the individual data points.
The macromolecular crowding agents (Dextran 40, Dextran 6 and
Ficoll 70) when added to a solution of α-synuclein (35 μM) decreased
the lag phase and increased the rate of formation of fibrils in a crowder
concentration dependent manner up to 100 g/L for all the three crowders
as shown by the ThT (Fig. 2) and CR assays (Fig. S2, Supporting Informa-
tion). The aggregation kinetics showed the characteristic sigmoidal
pattern for α-synuclein with a lag phase, log phase (elongation and for-
mation of protofibrils) and a final stationary phase analogous to the gen-
eration of mature amyloid fibrils (Fig. 2 and Fig. S2, Supporting
Information). The rate of aggregation of α-synuclein was slower in case
of Dextran 40 than in the presence of Dextran 6 and Ficoll 70 (Table 1a).
Addition of 150 g/L crowders resulted in an aggregation profile and rate
of fibril formation similar to that of 100 g/L crowders (Table 1a, Fig. 2
and Fig. S2, Supporting Information). It can be deduced that the maximum
fibril formation was observed at a concentration of 100 g/L for all the three
crowders. Beyond this concentration the trend was reversed, showing a
decrease in the rate of aggregation and fibril formation (Table 1a).
To further accelerate the aggregation kinetics of α-synuclein, etha-
nol was used. At low concentrations, ethanol enhances the rate of fibril
formation without changing the morphology of the fibrils [7]. Therefore,
ethanol at three different concentrations of 5% (v/v), 7.5% (v/v) and 10%
(v/v) was added to the reaction system having α-synuclein with 100 g/L
Dextran 40 (Fig. S3, Supporting Information). In our studies we used an
ethanol concentration of 7.5% (v/v) as the optimum concentration
both with regards to the rate of fibril formation and the ease with
which the aggregation process could be followed. Above this concen-
tration (10%, v/v) the solution turned turbid spontaneously, while at
a concentration below this (5%, v/v), the aggregation process was
quite slow (Table S1, Supporting Information).
Aggregation of α-synuclein in the presence of different concentrations
of macromolecular crowding agents both in absence (Fig. 2) and presence
of 7.5% (v/v) ethanol (Fig. S4, Supporting Information) again showed no
change in the pattern of aggregation kinetics. Moreover the highest
amount of fibril formation and fastest aggregation kinetics were observed
at a concentration of 100 g/L for all the three macromolecular crowding
agents irrespective of whether ethanol was present or not (Table 1a and
Table S2, Supporting Information). In addition, slower aggregation kinet-
ics along with increased lag time were observed with an increase in
crowder concentration beyond 150 g/L for both (in presence or absence
of ethanol). Thus the only alteration observed was a decrease in lag
time with faster aggregation for the samples having ethanol. These data
taken together demonstrate that the only effect of the organic solvent
was to increase the kinetics of aggregation and reduce the lag time keep-
ing the other trends/characteristics intact (Fig. S4, Supporting Informa-
tion). Finally, for the inhibition of aggregate formation and
disaggregation of preformed synuclein fibrils, our main focus was primar-
ily on the 100 g/L crowder concentration since the latter brought about
the maximum rate enhancement and fibril formation (Table 1a).
3.2. Inhibition of aggregation
First we studied the inhibition of aggregation of α-synuclein by indi-
vidual polyphenols. Addition of polyphenols decelerated the

Table 1
Kinetic parameters obtained (based on Eq. (1)) for aggregation and disaggregation (of preformed aggregates) of α-synuclein.

(a) Aggregation of α-synuclein in the presence of different concentrations of crowding agents (Fig. 2).

Concentration of the crowding agents (g/L) LOGt0⁎ for Dextran 40 LOGt0⁎ for Dextran 6 LOGt0⁎ for Ficoll 70
0 806.4 806.4 806.4
50 695.7 619.0 575.4
75 590.3 509.8 483.9
100 453.7 364.0 330.5
150 461.3 367.1 342.4
200 582.1 397.7 398.6
250 608.9 462.2 472.1
1.0
(b) Disaggregation of preformed aggregates of α-synuclein in the presence of three different macromolecular crowding agents (Dextran 40, Dextran 6 and Ficoll 70) at a
concentration of 100 g/L by the different optimized polyphenol–β-CD combinations (Fig. 5).

Crowding agent LOGt0⁎ for Curcumin and β-CD LOGt0⁎ for Resveratrol and β-CD LOGt0⁎ for Baicalein and β-CD LOGt0⁎ for EGCG and β-CD
Dextran 40 110.6 117.4 138.7 145.5
Dextran 6 131.0 136.2 159.6 163.9
Ficoll 70 132.8 138.9 162.5 165.9
⁎ LOGt0 = Time in min required for half maximum response.
594 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603

aggregation kinetics and increased the lag phase in a polyphenol con-
centration dependent manner (Fig. 3 and Fig. S5, Supporting Informa-
tion). The effective concentrations of polyphenols, where no
aggregation was observed, were 25 μM, 60 μM, 30 μM and 60 μM for
curcumin, resveratrol, baicalein and EGCG, respectively. These concen-
trations required for complete inhibition of aggregation of α-synuclein
were slightly higher than those necessary in the absence of any of the
macromolecular crowding agents (i.e. buffer only) (Table S3,
Supporting Information). The trend for effectiveness of polyphenols in
the presence of Dextran 40 was observed to be curcumin N baicalein N
resveratrol ≈ EGCG, as defined by the lowest concentrations of the re-
spective polyphenol required for complete inhibition. Addition of β-
CD alone to α-synuclein could inhibit the aggregation but only at a
much higher concentration of 200 μM (Fig. S6, Supporting Information).
This effect along with the known pseudochaperone action of β-CD [32]
and our previous results with curcumin [22] obligated us to use it in
combination with the polyphenols. Our earlier study has shown that ad-
dition of β-CD to curcumin in a ratio of 2:1 was most effective in
inhibiting and disaggregating α-synuclein amyloids [30]. In this work
too, the addition of β-CD along with polyphenols in a ratio of 2:1 in
the presence of Dextran 40 lowered the required concentration of
each of the polyphenols by a factor of 3 to 4 both in the absence of eth-
anol (Fig. 4 and Table S3, Supporting Information) as well as in its pres-
ence (Fig. S7, Supporting Information). The effective concentrations for
curcumin, resveratrol, baicalein and EGCG in combination with β-CD
were 7.5 μM (with 15 μM β-CD), 15 μM (with 30 μM β-CD), 10 μM
(with 20 μM β-CD) and 15 μM (with 30 μM β-CD), respectively. These
combinations were also effective with Dextran 6 (Fig. S8, Supporting In-
formation) and Ficoll 70 (Fig. S9, Supporting Information) both of which
induce faster aggregation than Dextran 40 (Fig. S10, Supporting Infor-
mation), thereby further proving the efficacy of the combinations as in-
hibitors. We also explored the effectiveness of these optimized
combinations for different crowder concentrations (0–200 g/L) and
found these to be able to completely abolish the aggregation of α-synu-
clein under these conditions (Fig. S11, Supporting Information).
3.3. Disaggregation
After successful inhibition of aggregation of α-synuclein by the
polyphenol–β-CD combinations, we extended our investigation to the
disaggregation of preformed amyloid fibrils of α-synuclein in the ab-
sence and presence of macromolecular crowding agents and ethanol.
The concentration of polyphenols and β-CD used were the same optimal
ones as obtained from the inhibition studies. Our results show that all
these combinations successfully disaggregated the preformed fibrils of
α-synuclein in the presence as well as absence of macromolecular
crowding agents (Fig. 5 and Fig. S12, Supporting Information). Irrespec-
tive of whether the aggregates are formed in the presence or absence of
ethanol and macromolecular crowding agents, the disaggregation pro-
files by the polyphenol–β-CD combination was similar in all cases
(Fig. S13, Supporting Information). The kinetics of disaggregation by
all the combinations demonstrated a sigmoidal curve similar to aggre-
gation kinetics after addition of the optimized combinations but with
less visible lag phase in all the cases (Fig. 5 and Fig. S12, Supporting In-
formation). The rate of disaggregation was significantly faster in buffer
only than in the presence of macromolecular crowding agents for all
the combinations, with the shortest time required being for the
curcumin–β-CD combination (Table 1b). Amongst the macromolecular
crowding agents, the fastest rate of disaggregation was observed for
Dextran 40 while Dextran 6 and Ficoll 70 were distinctly slower. All
the combinations were able to completely disaggregate preformed am-
yloids of α-synuclein even at the much higher concentration of 200 g/L
for crowders both with and without ethanol (Fig. S14, Supporting Infor-
mation). This, in combination with our previous observations, conclu-
sively proves that ethanol's role was only to enhance the rates of
aggregation. When an enhanced dose/concentration of the combination

Fig. 3. Effect of different polyphenols at various concentrations on the aggregation of α-synuclein using ThT fluorescence (λex = 430 nm and λem = 485 nm) in the presence of Dextran 40
(100 g/L) as the macromolecular crowding agent. The experiments were carried out at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing 0.1 M NaCl and 7.5% (v/v) ethanol with the
protein concentration being kept at 35 μM. The data plotted are a mean of the experiments done in triplicate and the standard deviation so obtained has been included for the individual
data points.
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603 595

Fig. 4. Effect of different polyphenols in combination with β-CD on the aggregation of α-synuclein using ThT fluorescence (λex = 430 nm and λem = 485 nm) in the presence of Dextran 40
(100 g/L) as the macromolecular crowding agent: (a) Curcumin with β-CD, (b) resveratrol with β-CD, (c) baicalein with β-CD and (d) EGCG with β-CD. The experiments were carried out
at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing 0.1 M NaCl with the protein concentration being maintained at 35 μM. The data plotted are a mean of the experiments done in
triplicate and the standard deviation so obtained has been included for the individual data points.

was used (keeping the polyphenol to β-CD ratio fixed at 1:2), the time again found to be the most effective one as observed from the rate of
required for disaggregation decreased (Fig. S15 and Table S4, disaggregation followed by baicalein, resveratrol and EGCG (Table S4,
Supporting Information). Curcumin in combination with β-CD was Supporting Information).

Fig. 5. Disaggregation of preformed aggregates of α-synuclein in the presence and absence of different macromolecular crowding agents: (a) curcumin (7.5 μM) with β-CD (15 μM), (b)
resveratrol (15 μM) with β-CD (30 μM), (c) baicalein (10 μM) with β-CD (20 μM) and (d) EGCG (15 μM) with β-CD (30 μM). Disaggregation profiles were measured using ThT fluorescence
(λex = 430 nm and λem = 485 nm) of preformed aggregates of α-synuclein (obtained after 800 min of aggregation) after addition of the respective polyphenol–β-CD combinations. The
experiments were carried out at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing 0.1 M NaCl with a protein concentration of 35 μM. The concentration of the macromolecular
crowding agents was 100 g/L. The data plotted are a mean of the experiments done in triplicate and the standard deviation so obtained has been included for the individual data points.
596 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603

3.4. Secondary structure of α-synuclein
Far UV circular dichroism spectroscopy was carried out to further
understand the effect of these polyphenols on the secondary structure
of α-synuclein. Native monomeric α-synuclein is an intrinsically disor-
dered protein and therefore showed a negative peak at around 200 nm
corresponding to its random structure (Fig. S16, Supporting Informa-
tion). Aggregation of α-synuclein by incubating in the presence of dif-
ferent macromolecular crowding agents resulted in the formation of
β-sheet rich structure which is a characteristic of amyloids (Fig. S16,
Supporting Information). Introduction of optimized concentrations of
different polyphenols along with β-CD maintained the native random
structure of α-synuclein thereby preventing the formation of any inter-
mediate molten globular or β-sheet conformation essential for amyloid
formation (Fig. S17a, Supporting Information). This supported the
above results obtained using ThT and CR. Furthermore, we added poly-
phenols in combination with β-CD to the preformed amyloid fibrils of
α-synuclein. The resultant α-synuclein species upon incubation with
polyphenols and β-CD reverted to the random native structure rather
than the initial β-sheet rich structure present in preformed amyloids
(Fig. S17b, Supporting Information).
3.5. Size exclusion chromatography (SEC) analyses
SEC analysis was carried out for the supernatants containing soluble
protein fractions after centrifugation. Elution profile of native mono-
meric α-synuclein during SEC analysis showed a peak resembling a
globular protein of 55 kDa (Fig. 6a). This is a well known property of
α-synuclein due to its intrinsically disordered nature and has been
shown by other groups too [56]. We then studied the aggregation of
α-synuclein using SEC as a function of time (Fig. 6a). After 20 min, the
intensity of peak corresponding to native monomeric α-synuclein
decreased considerably with simultaneous appearance of higher molec-
ular weight peaks, the latter signifying the presence of α-synuclein olig-
omers and protofibrils. At 60 min, the monomeric protein peak
disappeared completely and the peak due to intermediate oligomers
also started to disappear. This was accompanied by a decrease in the in-
tensity of the peak obtained near void volume corresponding to
protofibrils due to the formation of higher molecular weight and insol-
uble amyloid fibrils that cannot be analyzed using SEC. After 120 min,
there was only one peak corresponding to protofibrils with significantly
lower intensity. This decrease in the intensity of protofibrils was due to
the formation of higher molecular weight insoluble amyloid fibrils of α-
synuclein which were removed during centrifugation and filtration be-
fore SEC analysis.
We then investigated the effect of optimized combinations of poly-
phenols and β-CD on the elution profile of α-synuclein. No change
was observed in elution profile of α-synuclein (same as native α-synu-
clein) under aggregating conditions even after 120 min in the presence
of optimized combinations of polyphenols and β-CD namely: curcumin
(7.5 μM) with β-CD (15 μM), resveratrol (15 μM) with β-CD (30 μM),
baicalein (10 μM) with β-CD (20 μM) and EGCG (15 μM) with β-CD
(30 μM) (Fig. S18, Supporting Information). These data clearly demon-
strated that presence of these polyphenol–β-CD combinations did not
allow the protein to undergo aggregation or oligomerization and there-
by supported our ThT and CR data discussed earlier.
Disaggregation of the preformed aggregates of α-synuclein was also
analyzed using SEC (Fig. 6b). Addition of curcumin (7.5 μM) and β-CD
(15 μM) to the aggregated α-synuclein showed a trend reverse of that
observed during the aggregation process (Fig. 6a). After 30 min, elution
profile started to show peaks corresponding to oligomers and mono-
mers. The intensity of the monomeric α-synuclein increased with
time and only monomeric α-synuclein was present after 240 min show-
ing the completion of disaggregation process. Similar results were

Fig. 6. SEC analyses showing the elution profile of α-synuclein. (a) Aggregation of α-synuclein as a function of time and (b) Disaggregation profile after addition of curcumin (7.5 μM) with
β-CD (15 μM) to the preformed aggregates of α-synuclein. Aggregation was carried out at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing 0.1 M NaCl and ethanol (7.5%, v/v) with a
protein concentration of 35 μM in the presence of Dextran 40 (100 g/L). Disaggregation of preformed aggregates (obtained after 120 min of aggregation) was carried out after addition of
curcumin with β-CD under the same conditions. The arrows indicate the elution volumes of proteins used as molecular weight standards: (1) Void volume, (2) IgG (158 kDa), (3)
conalbumin (75 kDa), (4) bovine serum albumin (66.5 kDa), (5) ovalbumin (44 kDa), (6) super oxide dismutase 1 (32 kDa) and (7) lysozyme (14.4 kDa).
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603 597

obtained with other optimized combinations of resveratrol (15 μM)
with β-CD (30 μM), baicalein (10 μM) with β-CD (20 μM) and EGCG
(15 μM) with β-CD (30 μM) and disaggregation was completed within
300 min in all these cases (Fig. S19, Supporting Information) as was
also shown by the ThT and CR assays. The SEC data thus also pointed to-
wards the fact that these optimized combinations can not only inhibit
but also reverse the aggregation process.
3.6. Transmission electron microscopy (TEM)
Aggregation of α-synuclein with different macromolecular
crowding agents (Dextran 40, Dextran 6 and Ficoll 70) resulted in the
formation of amyloid fibrils with similar morphology (Fig. S20a&b,
Supporting Information). Incorporation of optimized combination of
different polyphenols with β-CD to the solution of α-synuclein in the
presence of macromolecular crowding agents completely inhibited the
formation of amyloid fibrils (Fig. S20d–g, Supporting Information). We
further analyzed the effect of these combinations of polyphenols and
β-CD on the preformed amyloid fibrils of α-synuclein (Fig. 7). All the
combinations disaggregated the preformed fibrils (Fig. 7 and Fig. S21,
Supporting Information) within the time duration (300 min) as ob-
served earlier using ThT fluorescence and CR absorbance. Therefore,
TEM images further confirmed the effectiveness of the unique, comple-
mentary/synergistic and potent combination of β-CD with various
polyphenols at sub-stoichiometric (with respect to that of α-synuclein)
concentrations.
3.7. Role of β-cyclodextrin (β-CD)
We also attempted to assess the role of β-CD used in combination
with the polyphenols in this study. As shown above, polyphenols
alone without β-CD were required at much higher concentration (Fig.
3 and Table S3, Supporting Information) than in combination to
completely inhibit the aggregation of α-synuclein. Substitution of β-
CD in the optimized combination by other commonly used cyclodextrin
variants had a negative influence on the inhibition of aggregation (Fig.
8a&b). Incorporation of either α-CD (Fig. 8a) or HP-β-CD [(2-Hydroxy-
propyl)-β-cyclodextrin] (Fig. 8b) along with the polyphenols (with the
same optimized concentrations as used earlier with β-CD) could not in-
hibit the aggregate formation by α-synuclein. This implied that neither
α-CD nor HP-β-CD were as efficient as β-CD. The effect of CDs only, that
is, without the presence of the polyphenolic compounds, on inhibition
of aggregation was also investigated (Fig. 8c). β-CD was effective in
inhibiting the aggregation only when used at a high concentration of
200 μM with α-CD being the least effective while HP-β-CD showed mar-
ginal insignificant inhibition (Fig. 8c). Thus further experiments with α-
CD or HP-β-CD on their abilities (in combination with the polyphenols)
to disaggregate α-synuclein aggregates were not performed. These data

Fig. 7. TEM images showing the disaggregation of preformed aggregates of α-synuclein in the presence of curcumin (7.5 μM) with β-CD (15 μM) at different time intervals of (a) 0 min, (b)
60 min, (c) 120 min, (d) 180 min (e) 210 min and (f) 240 min. Disaggregation of preformed aggregates of α-synuclein (obtained after 120 min of aggregation) was carried out after
addition of respective polyphenols with β-CD. Aggregation and disaggregation experiments were carried out at 37 °C with stirring in 10 mM Tris-HCl, pH 7.4 containing 0.1 M NaCl
and ethanol (7.5%, v/v) with a protein concentration of 35 μM in the presence of Dextran 40 (100 g/L). Scale bar is 200 nm.
598 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603

therefore established the important role played by β-CD and its essence
in the combination with polyphenols.

3.8. Stoichiometry between protein and polyphenols/β-CD combination is

also important

The optimized combinations of polyphenol and β-CD were able to

inhibit the aggregation of α-synuclein present at a concentration of 35
μM (Fig. 4). But when the concentration of the protein was increased
even by 5 μM (i.e. a total protein concentration of 40 μM), complete in-
hibition of aggregation was not achieved by the optimized combination
of curcumin (7.5 μM) and β-CD (15 μM) (Fig. S22a, Supporting Informa-
tion). Further increase in the concentration of α-synuclein (studied up
to 100 μM) enhanced the level of aggregation and was proportional to
an increase in protein concentration (Fig. S22a, Supporting Information)
as expected. It is worth noting that the ratio of α-synuclein (with 35 μM
concentration) to curcumin in the optimized combination (7.5 μM
curcumin along with 15 μM β-CD) is around 4.66. With increased con-
centration of α-synuclein (100 μM), similar protein to polyphenol
ratio of 5.0 was able to completely inhibit the aggregation (20 μM
curcumin along with 40 μM β-CD) (Fig. S22b, Supporting Information)
thereby implying the significance of the stoichiometry in the present
case.

3.9. Cytotoxicity assays

To investigate whether the aggregation of α-synuclein and the com-

Fig. 8. Effect of different cyclodextrins and polyphenols on the aggregation of α-synuclein
using ThT fluorescence (λex = 430 nm and λem = 485 nm). (a) α-CD in combination with
different polyphenols, (b) HP-β-CD in combination with different polyphenols and (c)
cyclodextrins only. The experiments were carried out at 37 °C with stirring in 10 mM
Tris-HCl, pH 7.4 containing 0.1 M NaCl and ethanol (7.5%, v/v) with a protein
concentration of 35 μM in the presence of Dextran 40 (100 g/L). [Abbreviations used:
cur = curcumin; res = resveratrol; baic = baicalein; EGCG = (−)-epigallocatechin
gallate; ACD = α-CD; HPBCD = 2-Hydroxypropyl-β-cyclodextrin and BCD = β-CD].
bination strategy employed by us for its inhibition or disaggregation can
be promoted as effective therapeutics, we tested their effect on the via-
bility of cultured N2a cells (mouse neuroblastoma cell lines). Since
prefibrillar aggregates are considered to be the most toxic species,
therefore cell viability assay was carried out with α-synuclein samples
taken at time points corresponding to prefibrillar aggregates (obtained
after 50 min of aggregation) [57]. As expected, α-synuclein at
prefibrillar stages induced drastic reduction (b 30%) in cell viability as
compared to the controls that had no α-synuclein (Fig. 9). However,
α-synuclein species incubated with polyphenols alone showed signifi-
cant (N80%) increase in cell viability as compared to the prefibrillar spe-
cies added without them (Fig. 9). Notably, when β-CD was also included
with these polyphenols a further increase in cell viability was observed,
with the most pronounced effect visible in the case of EGCG. Since the

Fig. 9. MTT reduction assay after 12 h incubation of N2a cells showing percentage viability in the presence or absence of α-synuclein prefibrillar aggregates (α-synuclein (P)). These are
species formed at 50 min time point of the aggregation kinetics. Cells incubated without or with polyphenols (curcumin, resveratrol, baicalein & ECGC) and/or β-CD for 12 h served as
positive controls, as represented by colored bars where there was almost no loss of viability. Incubation of cells with α-synuclein P alone reduced its viability drastically (to ~20%) and
hence served as negative control. Note that α-synuclein + β-CD combination also reduced cell viability. Compared to these, significant increase in cell viability (N80%) was observed
when cells were incubated with α-synuclein species formed in the presence of polyphenols and polyphenol + β-CD. P *p 0.05, non significant p values are not shown here.
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603 599

* *

Fig. 10. Effect of decreasing concentrations of curcumin in presence of fixed concentration of β-CD in molar ratios of 5:1, 5:0.75, 5:0.5, 5:0.25 (as shown by dark green color bars) as plotted
along with their respective controls (as shown by light green color bars) that shows significant increase in the cell protective activity of polyphenol (curcumin) in presence of β-CD. P *p
0.05, non significant p values are not shown here.

effect of β-CD was not that evident as the polyphenols themselves
showed substantial effects, we performed another series of experiments
where the concentration of curcumin, was lowered from its optimum
value of 7.5 μM (Fig. 10). The data reveal that there is a significant in-
crease in cell survivability when β-CD is used in combination with poly-
phenol, confirming our hypothesis on the synergism exhibited by the
polyphenol-βCD combiation. The most pronounced increase was ob-
served at the lowest concentration of curcumin (1.875 μM) in presence
of β-CD (37.5 μM) as compared to that of curcumin (1.875 μM) alone.
This effect however was visible only upto a certain stoichiometric
ratio of the polyphenol-βCD combination, beyond which the synergistic
effect showed a decrease (Fig. S23). This could be due to a narrow range
of operational stoichiometry where the disaggregation effect of poly-
phenol is enhanced by the presence of β-CD.
4. Molecular dynamics simulations
To gain atomistic insights into the mechanism of action of these
polyphenols, all-atom MD simulations were carried for α-synuclein
alone and in the presence of phenolic compounds. To correlate with
ex-vivo outcomes, solution structure of human micelle-bound α-synu-
clein was used. The amino acid sequence of α-synuclein can be divided

Fig. 11. Gibbs free energy landscapes projected as function RMSD (nm) and Rg (nm) for simulations of alpha-synuclein alone (A) and in the presence of curcumin, baicalein, resveratrol and
EGCG (B–E). Conformations corresponding to lowest free energy basin attained during simulations are shown in above panel. The corresponding lower panel indicates changes in
secondary structure contents over the entire simulation of 80 ns. Transitions of residues to beta-conformation in simulations of alpha-synuclein alone are shown as marked by the red
rectangle.
600 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
into three regions, N-terminal region comprising of residues 1–60, cen-
tral hydrophobic region consisting of residues 61–95 that are hydropho-
bic and C-terminal region comprising of residues 96–140, rich in acidic
residues (Fig. S1). The central region has proved to be crucial for aggre-
gation, as its truncation results in reduced aggregation of protein. The
Gibbs free energy landscape was projected as a function of backbone
RMSD and gyration radius (Rg). Backbone RMSD (root mean square de-
viation) is a measure of protein main-chain atomic deviations over the
entire trajectory of simulation as compared to its initial starting confor-
mation (called the reference frame). The radius of gyration (Rg) is the
measure of protein compactness by calculating gyration about the x, y
and z-axes, as a function of time with atoms considered explicitly
mass weighted. In simulation of α-synuclein alone, two distinct lowest
energy basins were observed (Fig. 11A). The corresponding coordinates
showed presence of distinct out of register/anti-parallel β-sheets
(shown in snapshots and secondary structure profiles of residues).
However, in the presence of the polyphenols, no such intra-molecular
rearrangements were observed and the low energy basin drifted in
the presence of compounds (Fig. 11B–E) towards a conformational sub-
space having higher Rg values, suggestive of higher disordered struc-
tures. Specifically in the presence of curcumin and EGCG, the lowest
energy basin drifted the most. Resveratrol however, instead of showing
such a drift, gave rise to multiple low energy basins accompanied by an
increase in the scatter of the Rg, signifying more conformational disor-
der. When analyzed, the corresponding lowest energy conformations
showed binding of all the polyphenols to the core central region resi-
dues. In the absence of compounds, these regions showed lower fluctu-
ations, possibly due to stable intra-molecular interaction networks.
Conversely, in the presence of the polyphenols, the core-amyloid re-
gions (47–56 and 68–78) of the proteins showed higher RMSF (root
mean square fluctuations) possibly due to frequent contacts with
these inhibitory compounds (Fig. S24). Moreover, these interactions
are dominated by stable H-bond networks (Fig. S25). A recent report
on the role of DNA intercalators in modulating amyloidogenesis process
show dominant role of hetero-aromatic associations [58,59]. Detailed
structural analysis revealed that besides other aromatic interactions,
the Phe-95 of the central hydrophobic region of α synuclein is involved
in aromatic associations with the planar groups of the compounds. Res-
veratrol was an exception showing more intricate aromatic interactions
involving Tyr 125, 133 and 136 of α synuclein (Fig. S26). The simulation
outcomes suggests that direct interactions of compounds through H-
bonding and aromatic associations subject monomeric α synuclein to
unstructured aggregates thus controlling fibrillogenesis and associated
cytotoxicity.
5. Discussion
The physiological interior of cells present a very complex environ-
ment, one which is not only congested, but at the same time also ex-
hibits a multitude of interactions (electrostatic, hydrophobic,
dispersion forces, excluded volume etc.) of the surrounding macromol-
ecules with the test protein [60]. Thus it is not surprising that dilute
buffer solutions might not be able to always provide an apt representa-
tion of the cellular interior. Moreover given the fact that protein aggre-
gation is itself a very heterogeneous and intricate process, this in
combination with the added vagaries of the crowded milieu make the
study of such self-assembly processes even more challenging. Further,
any related attempts at trying to inhibit aggregate formation and/or try-
ing to break up preformed protein aggregates need to be performed in
cellular-like environments such that any obvious shortcomings or prob-
lems can be addressed before taking such efforts to the next level (that
is, in vivo). In the present study, the aggregation of α-synuclein was ob-
served to be much accelerated in presence of the macromolecular
crowders as compared to that in buffer only (Table 1a), very similar to
what has been observed before [61]. Such an increase has been attribut-
ed to the excluded volume effect that these crowders are known to
exhibit [40]. Remarkably, at higher concentrations (N150 g/L) of the
macromolecular crowding agents, the rate of aggregation slowed
down, that is, showed a reverse trend for all the crowders (Table 1a).
Since the aggregation process includes the diffusional encounter of
monomers/oligomers with each other, it is natural to expect that the in-
crease of viscosity that accompanies such high crowder concentrations,
would slow down the self-assembly process thereby giving rise to the
observed reduction in rates [62]. However, it has also been shown that
viscosity is not the major or sole reason for the observed decrease in ag-
gregation kinetics [63,64]. Schreiber and coworkers [59] have discussed
four distinct regimes of behavior depending on the polymer concentra-
tion, namely, dilute, crossover, semi-dilute and concentrated regimes,
with context to protein-protein association rates. In the semi-dilute re-
gime, polymer concentration is high enough such that macromolecular
crowders form a cage-like environment around the associating protein
molecules thereby increasing the effective encounter rate leading to en-
hanced aggregation kinetics. However, in the concentrated regime, the
polymers have the tendency to be situated in between the aggregating
species, posing as obstacles and thereby slowing down the assembly
process by minimizing complex formation. Our observations are a
clear reflection of this concept with 100–150 g/L corresponding to the
semi-dilute regime while beyond 200 g/L the concentrated polymer re-
gime comes into effect. On the other hand, the difference in the afore-
mentioned rates of fibril formation in the presence of different
macromolecular crowding agents (Table 1a) are attributed to the intrin-
sic differences in the molecular nature of the polymeric crowder mole-
cules exerting different extents of excluded volume. While dextran is a
polymer of glucose, Ficoll 70 is formed from the copolymerization of su-
crose with epichlorohydrin (Fig. 1). It is evident that both these crowder
molecules (Fig. 1) have enough \\OH groups to exhibit polar interac-
tions and\\CH2\\moieties for hydrophobic contacts with the test pro-
tein. Such soft interactions between protein and crowder molecules
have been shown to play an important role in protein conformational
distribution [65]. Dextran 6 with lowest average molecular weight
among the crowders used here has the maximum number of molecules
present at any given concentration giving rise to a higher packing den-
sity. This results in an increased excluded volume effect and hence the
faster aggregation rates for Dextran 6 as compared to Dextran 40 was
expected. Had one followed the same line of argument, Ficoll 70 should
have shown the least effect on the aggregation. However, our observa-
tions reveal that both Dextran 6 and Ficoll 70 have comparable rates
suggesting that excluded volume is not the only dominant factor. We
propose that the enhancement of aggregation rates seen in presence
of Ficoll 70, can be attributed to the presence of significant soft interac-
tions between protein and crowder molecules in combination with the
excluded volume so exerted at the concentrations of the crowding agent
used. Thus, the slower aggregation kinetics seen in presence of Dextran
40 arises from the fact that this being of intermediate molecular weight,
it neither exerts as high an excluded volume as Dextran 6 nor exhibits
the extent of soft interactions as that of Ficoll 70.
Based on the complexity of a congested environment, it is expected
that the concentration of polyphenols and β-CD required for complete
inhibition of aggregation will be different than those in absence of the
macromolecular crowding agents. Indeed this is apparent by the fact
that the polyphenols alone or in combination with β-CD were required
in a slightly higher concentration to bring about complete inhibition of
aggregation in presence of the macromolecular crowders (Table S3,
Supporting Information). Curcumin in combination with β-CD was
most effective with lowest concentration required (among all four poly-
phenols used) for inhibiting and disaggregating amyloids formed by α-
synuclein. It was followed by baicalein while resveratrol and EGCG were
equally effective in combination with β-CD.
The next obvious question is ‘why β-CD’? Our comparative studies
with other commonly used cyclodextrin variants revealed that β-CD is
the most effective one. Among the naturally occurring cyclodextrins,
β-CD contains seven glucose units as compared to six in case of α-CD
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
and eight for γ-CD which directly correlates with the size of internal hy-
drophobic cavity being 4.7–5.3 Å, 6.0–6.5 Å and 7.5–8.3 Å for α-CD, β-
CD and γ-CD, respectively [30]. β-CD, not being on the extreme ends
in terms of size and properties, is the most commonly used cyclodextrin
in pharmaceutical formulations and drug delivery [34]. In the current
study, β-CD showed excellent synergism along with the polyphenols
in not only helping in inhibition of the aggregation but also disaggregat-
ing fibrils at sub-stoichiometric concentrations. Moreover, β-CD helps
in overcoming the limitations of polyphenolic compounds such as low
bioavailability, in vivo instability [37], low efficiency [30] and inability
to cross blood-brain barrier at higher concentrations (i.e. 20 μM) etc.
[66]. Independent studies from our group have shown that β-CD signif-
icantly enhanced the in vitro stability of curcumin as compared to α-CD
or HP-β-CD. HP-β-CD, a hydroxypropyl derivative of β-CD with similar
cavity dimensions but enhanced water solubility, was however not very
effective and helpful for inhibiting amyloid formation. Thus, it can be
deduced that the optimum hydrophobicity (such as that present in β-
CD) along with the cavity size is an important pre-requisite for
exhibiting the effects as seen in this study.
The inhibition of α-synuclein aggregation involves the synergistic
effect of both the polyphenols and β-CD. Curcumin has been shown to
act by concealing the hydrophobic patches on the protein molecules
and later on interacting via hydrogen bonding with the side chains of
amino acid residues present in protein molecules [24,29]. This interferes
with two most important pre-requisites for amyloid formation: initial
hydrophobic interactions required for nucleation and its growth and
second, the hydrogen bonds stabilizing the β-sheets of amyloids [67].
The two aromatic rings impart curcumin the ability to exhibit hydro-
phobic interactions while the phenolic\\OH groups along with the car-
bonyl groups help in interacting with protein molecules via hydrogen
bonds. The linker between both the phenolic rings along with the ability
for keto-enol tautomerism provides curcumin with a greater flexibility
in terms of adapting conformations that can help facilitate interactions
leading to the impediment of protein aggregation. Moreover, Reinke
and Gestwicki [68] have proposed three important features in mole-
cules required for inhibiting fibrillation: (a) existence of two terminal
aromatic groups required for interacting with protein molecules, (b)
substitution on the aromatic groups i.e. presence of hydroxyl groups
and (c) the length of the linker connecting the two aromatic groups
which should be 6–19 Å. Curcumin fulfills all these criteria and it is
therefore not surprising that this polyphenol is the most effective
among the ones studied here. A recent study has also provided insights
into the mechanism of action of curcumin and similarly found that it in-
teracts via both hydrophobic effects and hydrogen bonding [69]. Pres-
ence of β-CD further helps in reducing the availability of hydrophobic
patches by interacting with these, especially with the phenylalanine
residues present in α-synuclein [36]. The three other polyphenols (res-
veratrol, baicalein and EGCG) seem to interact mainly through the –OH
groups by forming hydrogen bonds with residues of α-synuclein due to
the presence of vicinal polyhydroxyphenyl groups [10,13,14,70]. There-
after, hydrophobic interactions take place and minimize further inter-
molecular interactions among α-synuclein molecules [71]. EGCG con-
tains more\\OH groups than other polyphenols used in this study but
is least effective and is required at the highest concentration for
inhibiting or disaggregating amyloids. It must be emphasized that the
size of EGCG arising from the bulkiness of three aromatic rings is higher
than others. This results in sterically limiting the movement and flexibil-
ity of side groups present in EGCG as compared to curcumin which con-
sists of a long aliphatic linker between two aromatic rings. On the
contrary, though resveratrol contains a small aliphatic linker, it lacks
the vicinal\\OH groups which along with the restricted rotation around
its aliphatic double bond makes it far less effective than curcumin and
baicalein. Baicalein meanwhile possesses three vicinal polyhydroxy
groups which coupled with the smaller size of the molecule, help in
interacting even with interior i.e. relatively inaccessible segments of
the aggregates. Previous studies have emphasized the presence of
601
vicinal \\OH groups as being one of the primary characteristics of the
polyphenols [10,13] that are able to inhibit aggregation of α-synuclein.
However, our results reveal that other factors such as molecular flexibil-
ity, hydrophobic interactions and size, as discussed above, also play a
major role in inhibition or disaggregation of protein aggregates. In this
regard, curcumin, which does not have vicinal polyhydroxy groups
but still is the most effective among the compounds studied here, is an
apt example of the possible structural aspects that an effective com-
pound needs to have when so designed.
The polyphenol–β-CD combinations were also effective in complete-
ly disaggregating the preformed aggregates of α-synuclein. Surprising-
ly, the rate of disaggregation in the presence of Dextran 40 was faster as
compared to Dextran 6 or Ficoll 70, this being just opposite to the trend
observed for the aggregation kinetics (Table 1b and Fig. 5). This observa-
tion fits quite well into our following hypothesis: since Dextran 40 nei-
ther exhibits high excluded volume nor extensive soft interactions, the
access of the combination to the aggregates is easier thereby leading
to faster disassembly. Curcumin–β-CD showed the fastest disaggrega-
tion kinetics than the other combinations in the presence of all three
macromolecular crowding agents (Table 1b). Dose dependent increase
in the rate of disaggregation (Table S4 and Fig. S15, Supporting Informa-
tion) further highlighted that the combination is working in a synergis-
tic way with curcumin being most effective. The efficacy of the
combinations used for disaggregation was same as that observed for
the inhibition of aggregate formation, thereby suggesting that similar
types of interactions between the polyphenol–β-CD units and the pro-
teins were present. Berhanu & Masunov et al. [29,32] have shown
using MD simulations that polyphenols such as curcumin, exifone and
myricetin disaggregate the fibrils by forming hydrogen bonds not just
with side chain \\NH groups of protein molecules in fibrils but with
the \\NH groups present in peptide backbone as well. This results in
the loss of fibrillar hydrogen bonding leading to the dissociation of β-
sheets. Besides the intermolecular hydrogen bonding that the polyphe-
nols used in this study can exhibit with the aggregates, the presence of
β-CD ensures, through hydrophobic interactions, that on dissociation,
the component units do not recombine, thereby preventing auxiliary
aggregation [72]. Besides, EGCG has been shown to exhibit a range of in-
teractions with monomeric and oligomeric proteins thereby hinting at
the universality of these polyphenols being used more aggressively in
developing effective therapeutic approaches for neurodegenerative dis-
eases arising from protein misfolding and aggregation [14,70].
There exist some key differences between the mechanisms of inhibi-
tion of aggregation and disaggregation. First, while for inhibition the
concentration of polyphenols (7.5 μM for curcumin, 10 μM for baicalein
and 15 μM for both resveratrol and EGCG) used is lesser than that of α-
synuclein (35 μM), however during disaggregation the scenario is re-
versed. In the latter, presence of fibrils or higher molecular weight ag-
gregates ensures that their concentration is much reduced as
compared to the prevailing concentration of the polyphenols, since
each aggregate is composed of many synuclein monomers. Second,
the nature of the surface of α-synuclein exposed for interactions with
the polyphenols is different for the processes of inhibition and disaggre-
gation. During inhibition, the exposed hydrophobic stretch (NAC re-
gion) of the disordered protein, which is the main nucleation point for
association, is blocked by the combination(s), presumably initiated via
interactions that are predominantly hydrophobic. On the other hand,
in the aggregates, along with the presence of extensively hydrogen
bonded β-sheets well-defined hydrophobic clefts also exist amongst
the interdigitized monomers/oligomers. Such voids/cavities provide
more extensive hydrophobic interactions with the polyphenols used,
thereby further abetting subsequent cleavage into monomers with β-
CD helping to sequester the dissociated components through its
pseudochaperone-like action. Molecular dynamics simulations provid-
ed important insights on association of the compounds with monomeric
α-synuclein through favourable H-bond and aromatic interactions.
These interactions altered the conformational free energy subspace of
602 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
α-synuclein restricting its coil to cross-β transitions. Absence of such
transitions by polyphenols helped in remodelling the protein into an al-
ternative state which is less prone to aggregation. Such transitions could
be the triggering point for full amyloidogenesis of the protein.
Interestingly the prefibrillar aggregates of α-synuclein in presence
of β-CD along with polyphenolic compounds were much less toxic
than prefibrils alone as shown by cytotoxicity studies with N2a cells.
However, the polyphenols alone were also effective in significantly re-
ducing the toxicity of α-synuclein prefibrils. Polyphenols alone have
been shown to reduce the toxicity of α-synuclein oligomers/fibrils ear-
lier as well [11,14]. Recent studies based on the interaction of EGCG with
amyloids have proposed that this naturally occurring polyphenol brings
about restructuring of amyloids thereby rendering these aggregates
non-toxic and hence increasing cell-viability [14,71]. Otzen and co-
workers [73] have reported that EGCG helps in reducing cell toxicity in-
duced by α-synuclein oligomers by preventing membrane disruption
rather than bringing about any modulation of the structures of the pro-
tein aggregates. Reducing the concentration of curcumin by keeping
that of β-CD fixed (Fig. 10) clearly brought out the effect of this pseu-
do-chaperone and hence the role it plays in making this combination
a potent one. These and many other studies, including the present
one, thus suggest that while polyphenols provide a promising approach
for targeting neurodegenerative diseases, however, the exact mecha-
nisms of their actions, in combination with improved bioavailability of
the same [74] need to be explored in depth at the molecular level before
taking the next step towards drug design/formulation.
6. Conclusion
The combination(s) of the drug encapsulator β-CD and four poly-
phenols are effective not only in inhibiting α-synuclein aggregation
but also in breaking up preformed aggregates of the disordered protein,
at much lower doses than had the components been used in isolation.
The different chemical structures of these polyphenols have allowed
us to throw insights into the mechanistic aspects underlining the
mode of action of the polyphenol–β-CD combinations. While the effica-
cy of a polyphenol depends on a host of factors like structural flexibility,
presence of phenolic \\OH groups and a flavour of hydrophobicity as
brought in by the benzene moiety, the uniqueness of β-CD arises from
its apt cavity size and associated hydrophobic interior thereby allowing
the latter to interact with hydrophobic amino acid side chains and
hence block subsequent self-association of α-synuclein. Moreover, the
ability of β-CD to help solubilize and stabilize the polyphenols through
known hydrophobic encapsulation thus making these more bioavail-
able is an added advantage point of the combination employed in this
study. Macromolecular crowding influenced the aggregation kinetics
as expected but in a subtle manner, such that the effects observed
could not be ascribed to the excluded volume phenomenon alone. Addi-
tionally, the fact that the sub-stoichiometric concentrations of these
compounds were equally effective under cell-mimicking conditions fur-
ther lend credence to and provide a scope for developing these with an
eye on possible therapeutics. The ex vivo experiments with N2a cells
(mouse neuroblastoma cells) also supported the in vitro results demon-
strating the effectiveness of the four polyphenols and their concoction
with β-CD. This successful correlation exemplified the significance of
using cell mimicking conditions (i.e. macromolecular crowding agents)
while carrying out experiments in vitro rather than dilute buffer solu-
tions. Since curcumin and EGCG have already received attention in
modulating the aggregation of several neurodegenerative disease caus-
ing proteins, it is worthwhile to try the aforementioned polyphenol–β-
CD cocktail(s) for such cases too in order to probe the universality of the
prescribed approach. Indeed with extensive focus worldwide on com-
bating neurodegeneration and related ailments, such examples as
used in this study provide alternate promising routes that one can em-
bark upon, involving more focused attention and subsequent research,
to delineate the details about the underlying forces at play at the molec-
ular level.
Transparency document
The Transparency document associated with this article can be
found, in the online version.
Acknowledgements
PKC thanks the Department of Science and Technology (DST), New
Delhi, India, for financial support under the Fast Track Scheme for Young
Scientists (SR/FT/CS-007/2010) and IIT Delhi for startup funding. SG, PS
and PP thank CSIR and SK thanks IIT Delhi for providing their fellow-
ships. The authors thank IIT Delhi HPC facility for computational
resources.
References
[1] D. Eisenberg, M. Jucker, The amyloid state of proteins in human diseases, Cell 148
(2012) 1188–1203.
[2] O. Marques, T.F. Outeiro, Alpha-synuclein: from secretion to dysfunction and death,
Cell Death Dis. 3 (2012), e350.
[3] M. Bisaglia, S. Mammi, L. Bubacco, Structural insights on physiological functions and
pathological effects of alpha-synuclein, FASEB J. 23 (2009) 329–340.
[4] F. Cheng, G. Vivacqua, S. Yu, The role of alpha-synuclein in neurotransmission and
synaptic plasticity, J. Chem. Neuroanat. 42 (2011) 242–248.
[5] V.N. Uversky, Neuropathology, biochemistry, and biophysics of alpha-synuclein ag-
gregation, J. Neurochem. 103 (2007) 17–37.
[6] J.T. Bendor, T.P. Logan, R.H. Edwards, The function of alpha-synuclein, Neuron 79
(2013) 1044–1066.
[7] L.A. Munishkina, C. Phelan, V.N. Uversky, A.L. Fink, Conformational behavior and ag-
gregation of alpha-synuclein in organic solvents: modeling the effects of mem-
branes, Biochemistry 42 (2003) 2720–2730.
[8] G. Toth, S.J. Gardai, W. Zago, C.W. Bertoncini, N. Cremades, S.L. Roy, M.A. Tambe, J.C.
Rochet, C. Galvagnion, G. Skibinski, S. Finkbeiner, M. Bova, K. Regnstrom, S.S. Chiou,
J. Johnston, K. Callaway, J.P. Anderson, M.F. Jobling, A.K. Buell, T.A. Yednock, T.P.
Knowles, M. Vendruscolo, J. Christodoulou, C.M. Dobson, D. Schenk, L.
McConlogue, Targeting the intrinsically disordered structural ensemble of alpha-sy-
nuclein by small molecules as a potential therapeutic strategy for Parkinson's dis-
ease, PLoS One 9 (2014) e87133.
[9] L. Breydo, J.W. Wu, V.N. Uversky, Alpha-synuclein misfolding and Parkinson's dis-
ease, Biochim. Biophys. Acta 1822 (2012) 261–285.
[10] M. Caruana, T. Högen, J. Levin, A. Hillmer, A. Giese, N. Vassallo, Inhibition and disag-
gregation of α-synuclein oligomers by natural polyphenolic compounds, FEBS Lett.
585 (2011) 1113–1120.
[11] P.K. Singh, V. Kotia, D. Ghosh, G.M. Mohite, A. Kumar, S.K. Maji, Curcumin modulates
α-synuclein aggregation and toxicity, ACS Chem. Neurosci. 4 (2013) 393–407.
[12] M.E. Herva, S. Zibaee, G. Fraser, R.A. Barker, M. Goedert, M.G. Spillantini, Anti-amy-
loid compounds inhibit alpha-synuclein aggregation induced by protein misfolding
cyclic amplification (PMCA), J. Biol. Chem. 289 (2014) 11897–11905.
[13] X. Meng, L.A. Munishkina, A.L. Fink, V.N. Uversky, Molecular mechanisms underlying
the flavonoid-induced inhibition of alpha-synuclein fibrillation, Biochemistry 48
(2009) 8206–8224.
[14] J. Bieschke, J. Russ, R.P. Friedrich, D.E. Ehrnhoefer, H. Wobst, K. Neugebauer, E.E.
Wanker, EGCG remodels mature alpha-synuclein and amyloid-beta fibrils and re-
duces cellular toxicity, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 7710–7715.
[15] K.J. Korshavn, M. Jang, Y.J. Kwak, A. Kochi, S. Vertuani, A. Bhunia, S. Manfredini, A.
Ramamoorthy, M.H. Lim, Reactivity of metal-free and metal-associated amyloid-β
with glycosylated polyphenols and their esterified derivatives, Sci. Rep. 5 (2015)
17842–17856.
[16] A.S. Detoma, S. Salamekh, A. Ramamoorthy, M.H. Lim, Misfolded proteins in
Alzheimer's disease and type II diabetes, Chem. Soc. Rev. 41 (2012) 608–621.
[17] S.S. Hindo, A.M. Mancino, J.J. Braymer, Y. Liu, S. Vivekanandan, A. Ramamoorthy,
M.H. Lim, Small molecule modulators of copper-induced Aβ aggregation, J. Am.
Chem. Soc. 131 (2009) 16663–16665.
[18] S. Lee, X. Zheng, J. Krishnamoorthy, M.G. Savelieff, H.M. Park, J.R. Brender, J.H. Kim,
J.S. Derrick, A. Kochi, H.J. Lee, C. Kim, A. Ramamoorthy, M.T. Bowers, M.H. Lim, Ratio-
nal design of a structural framework with potential use to develop chemical re-
agents that target and modulate multiple facets of Alzheimer's disease, J. Am.
Chem. Soc. 136 (2013) 299–310.
[19] S.J. Hyunga, A.S. DeTomaa, J.R. Brender, S. Lee, S. Vivekanandan, A. Kochia, J.S. Choic,
A. Ramamoorthya, B.T. Ruotoloa, M.H. Lim, Insights into antiamyloidogenic proper-
ties of the green tea extract (−)-epigallocatechin-3-gallate toward metal-associated
amyloid-β species, Proc. Natl. Acad. Sci. 110 (2013) 3743–3748.
[20] J.S. Choi, J.J. Braymer, R.P.R. Nanga, A. Ramamoorthy, M.H. Lim, Design of small mol-
ecules that target metal-Aβ species and regulate metal-induced Aβ aggregation and
neurotoxicity, Proc. Natl. Acad. Sci. 107 (2010) 21990–21995.
[21] W.M. Berhanu, A.E. Masunov, Natural polyphenols as inhibitors of amyloid aggrega-
tion. Molecular dynamics study of GNNQQNY heptapeptide decamer, Biophys.
Chem. 149 (2010) 12–21.
 S. Gautam et al. / Biochimica et Biophysica Acta 1865 (2017) 589–603
[22] S.A. Hudson, H. Ecroyd, F.C. Dehle, I.F. Musgrave, J.A. Carver, (−)-Epigallocatechin-3-
gallate (EGCG) maintains kappa-casein in its pre-fibrillar state without redirecting
its aggregation pathway, J. Mol. Biol. 392 (2009) 689–700.
[23] D. Vauzour, A. Rodriguez-Mateos, G. Corona, M.J. Oruna-Concha, J.P. Spencer, Poly-
phenols and human health: prevention of disease and mechanisms of action, Nutri-
ents 2 (2010) 1106–1131.
[24] W.M. Berhanu, A.E. Masunov, Atomistic mechanism of polyphenol amyloid aggrega-
tion inhibitors: molecular dynamics study of curcumin, exifone, and myricetin inter-
action with the segment of tau peptide oligomer, J. Biomol. Struct. Dyn. 1-13 (2014).
[25] B.B. Aggarwal, C. Sundaram, N. Malani, H. Ichikawa, Curcumin: the Indian solid gold,
Adv. Exp. Med. Biol. 595 (2007) 1–75.
[26] N. Pandey, J. Strider, W.C. Nolan, S.X. Yan, J.E. Galvin, Curcumin inhibits aggregation
of alpha-synuclein, Acta Neuropathol. 115 (2008) 479–489.
[27] A.S. Pithadia, A. Bhunia, R. Sribalan, V. Padmini, C.A. Fierke, A. Ramamoorthy, Influ-
ence of a curcumin derivative on hIAPP aggregation in the absence and presence
of lipid membranes, Chem. Commun. 52 (2016) 942–945.
[28] P. Arya, A. Srivastava, S.V. Vasaikar, G. Mukherjee, P. Mishra, B. Kundu, Selective in-
terception of gelsolin amyloidogenic stretch results in conformationally distinct ag-
gregates with reduced toxicity, ACS Chem. Neurosci. 5 (2014) 982–992.
[29] B. Ahmad, L.J. Lapidus, Curcumin prevents aggregation in α-synuclein by increasing
reconfiguration rate, J. Biol. Chem. 287 (2012) 9193–9199.
[30] S. Gautam, S. Karmakar, A. Bose, P.K. Chowdhury, beta-Cyclodextrin and curcumin, a
potent cocktail for disaggregating and/or inhibiting amyloids: a case study with
alpha-synuclein, Biochemistry 53 (2014) 4081–4083.
[31] E.M.M. Del Valle, Cyclodextrins and their uses: a review, Process Biochem. 39 (2004)
1033–1046.
[32] S. Machida, S. Ogawa, S. Xiaohua, T. Takaha, K. Fujii, K. Hayashi, Cycloamylose as an
efficient artificial chaperone for protein refolding, FEBS Lett. 486 (2000) 131–135.
[33] F. Hapiot, H. Bricout, S. Tilloy, E. Monflier, Functionalized cyclodextrins as first and
second coordination sphere ligands for aqueous organometallic catalysis, Eur. J.
Inorg. Chem. (2012) 1571–1578.
[34] J. Zhang, P.X. Ma, Cyclodextrin-based supramolecular systems for drug delivery: re-
cent progress and future perspective, Adv. Drug Deliv. Rev. 65 (2013) 1215–1233.
[35] Z.T. Wehrmann, T.W. Hulett, K.L. Huegel, K.T. Vaughan, O. Wiest, P. Helquist, H.
Goodson, Quantitative comparison of the efficacy of various compounds in lowering
intracellular cholesterol levels in Niemann-Pick type C fibroblasts, PLoS One 7
(2012), e48561.
[36] X.R. Qin, H. Abe, H. Nakanishi, NMR and CD studies on the interaction of Alzheimer
beta-amyloid peptide (12-28) with beta-cyclodextrin, Biochem. Biophys. Res.
Commun. 297 (2002) 1011–1015.
[37] Z.X. Fang, B. Bhandari, Encapsulation of polyphenols—a review, Trends Food Sci.
Technol. 21 (2010) 510–523.
[38] V.R. Yadav, S. Prasad, R. Kannappan, J. Ravindran, M.M. Chaturvedi, L. Vaahtera, J.
Parkkinen, B.B. Aggarwal, Cyclodextrin-complexed curcumin exhibits anti-inflam-
matory and antiproliferative activities superior to those of curcumin through higher
cellular uptake, Biochem. Pharmacol. 80 (2010) 1021–1032.
[39] M. D'Archivio, C. Filesi, R. Vari, B. Scazzocchio, R. Masella, Bioavailability of the poly-
phenols: status and controversies, Int. J. Mol. Sci. 11 (2010) 1321–1342.
[40] H.X. Zhou, G. Rivas, A.P. Minton, Macromolecular crowding and confinement: bio-
chemical, biophysical, and potential physiological consequences, Annu. Rev.
Biophys. 37 (2008) 375–397.
[41] D. Gnutt, M. Gao, O. Brylski, M. Heyden, S. Ebbinghaus, Excluded-volume effects in
living cells, Angew. Chem. Int. Ed. Engl. 54 (2015) 2548–2551.
[42] J. Baell, M.A. Walters, Chemistry: Chemical con artists foil drug discovery, Nature
513 (2014) 481–483.
[43] M.J. Volles, P.T. Lansbury, Relationships between the sequence of alpha-synuclein
and its membrane affinity, fibrillization propensity, and yeast toxicity, J. Mol. Biol.
366 (2007) 1510–1522.
[44] H. LeVine, Quantification of beta-sheet amyloid fibril structures with thioflavin T,
Methods Enzymol. 309 (1999) 274–284.
[45] W.E. Klunk, R.F. Jacob, R.P. Mason, Quantifying amyloid by congo red spectral shift
assay, Methods Enzymol. 309 (1999) 285–305.
[46] J. van Meerloo, G.J. Kaspers, J. Cloos, Cell sensitivity assays: the MTT assay, Methods
Mol. Biol. 731 (2011) 237–245.
[47] GROMACS: high performance molecular simulations through multi-level parallel-
ism from laptops to supercomputers.
[48] Z. Lin, W.F. van Gunsteren, Refinement of the application of the GROMOS 54A7 force
field to beta-peptides, J. Comput. Chem. 34 (2013) 2796–2805.
[49] K.B. Koziara, M. Stroet, A.K. Malde, A.E. Mark, Testing and validation of the Automat-
ed Topology Builder (ATB) version 2.0: prediction of hydration free enthalpies, J.
Comput. Aided Mol. Des. 28 (2014) 221–233.
603
[50] A.K. Malde, L. Zuo, M. Breeze, M. Stroet, D. Poger, P.C. Nair, C. Oostenbrink, A.E. Mark,
An automated force field topology builder (ATB) and repository: version 1.0, J.
Chem. Theory Comput. 7 (2011) 4026–4037.
[51] M. Parinello, A. Rahman, Polymorphic transitions in single crystals: a new molecular
dynamics method, J. Appl. Phys. 52 (1981) 7182–7190.
[52] T. Darden, D. York, L. Pedersen, Particle mesh Ewald: an N-log(N) method for Ewald
sums in large systems, J. Chem. Phys. 98 (1993) 10089–10092.
[53] LLC. Schrodinger, The PyMOL Molecular Graphics System, Version 1.8, 2015.
[54] W. Humphrey, A. Dalke, K. Schulten, VMD—visual molecular dynamics, J. Mol.
Graph. 14 (1996) 33–38.
[55] Y. Suzuki, J.R. Brender, K. Hartman, A. Ramamoorthy, E.N.G. Marsh, Alternative path-
ways of human islet amyloid polypeptide aggregation distinguished by 19F nuclear
magnetic resonance detected kinetics of monomer consumption, Biochemistry 51
(2012) 8154–8162.
[56] W. Ariesandi, C.F. Chang, T.E. Chen, Y.R. Chen, Temperature-dependent structural
changes of Parkinson's alpha-synuclein reveal the role of pre-existing oligomers in
alpha-synuclein fibrillization, PLoS One 8 (2013), e53487.
[57] D.P. Karpinar, M.B. Balija, S. Kugler, F. Opazo, N. Rezaei-Ghaleh, N. Wender, H.Y. Kim,
G. Taschenberger, B.H. Falkenburger, H. Heise, A. Kumar, D. Riedel, L. Fichtner, A.
Voigt, G.H. Braus, K. Giller, S. Becker, A. Herzig, M. Baldus, H. Jackle, S. Eimer, J.B.
Schulz, C. Griesinger, M. Zweckstetter, Pre-fibrillar alpha-synuclein variants with
impaired beta-structure increase neurotoxicity in Parkinson's disease models,
EMBO J. 28 (2009) 3256–3268.
[58] J. Singh, A. Srivastava, P. Sharma, P. Pradhan, B. Kundu, DNA intercalators as amyloid
assembly modulators: mechanistic insights, RSC Adv. 7 (2017) 493–506.
[59] A. Srivastava, S. Sharma, S. Sadanandan, S. Gupta, J. Singh, S. Gupta, V. Harids, B.
Kundu, Modulation of prion polymerization and toxicity by rationally designed
peptidomimetics, Biochem. J. 474 (2017) 123–147.
[60] R.J. Ellis, Macromolecular crowding: an important but neglected aspect of the intra-
cellular environment, Curr. Opin. Struct. Biol. 11 (2001) 114–119.
[61] V.N. Uversky, M.E. Cooper, K.S. Bower, J. Li, A.L. Fink, Accelerated alpha-synuclein fi-
brillation in crowded milieu, FEBS Lett. 515 (2002) 99–103.
[62] L.A. Munishkina, E.M. Cooper, V.N. Uversky, A.L. Fink, The effect of macromolecular
crowding on protein aggregation and amyloid fibril formation, J. Mol. Recognit. 17
(2004) 456–464.
[63] Y.Y. Kuttner, N. Kozer, E. Segal, G. Schreiber, G. Haran, Separating the contribution of
translational and rotational diffusion to protein association, J. Am. Chem. Soc. 127
(2005) 15138–15144.
[64] N. Kozer, Y.Y. Kuttner, G. Haran, G. Schreiber, Protein-protein association in polymer
solutions: from dilute to semidilute to concentrated, Biophys. J. 92 (2007)
2139–2149.
[65] L. Sapir, D. Harries, Is the depletion force entropic? Molecular crowding beyond ste-
ric interactions, Curr. Opin. Colloid Interface Sci. 20 (2015) 3–10.
[66] L. Baum, A. Ng, Curcumin interaction with copper and iron suggests one possible
mechanism of action in Alzheimer's disease animal models, J. Alzheimers Dis. 6
(2004) 367–377 (discussion 443-369).
[67] K.E. Marshall, K.L. Morris, D. Charlton, N. O'Reilly, L. Lewis, H. Walden, L.C. Serpell,
Hydrophobic, aromatic, and electrostatic interactions play a central role in amyloid
fibril formation and stability, Biochemistry 50 (2011) 2061–2071.
[68] A.A. Reinke, J.E. Gestwicki, Structure-activity relationships of amyloid beta-aggrega-
tion inhibitors based on curcumin: influence of linker length and flexibility, Chem.
Biol. Drug Des. 70 (2007) 206–215.
[69] N.K. Bhatia, A. Srivastava, N. Katyal, N. Jain, M.A. Khan, B. Kundu, S. Deep, Curcumin
binds to the pre-fibrillar aggregates of Cu/Zn superoxide dismutase (SOD1) and al-
ters its amyloidogenic pathway resulting in reduced cytotoxicity, Biochim. Biophys.
Acta 1854 (2015) 426–436.
[70] P. Cao, D.P. Raleigh, Analysis of the inhibition and remodeling of islet amyloid poly-
peptide amyloid fibers by flavanols, Biochemistry 51 (2012) 2670–2683.
[71] F.L. Palhano, J. Lee, N.P. Grimster, J.W. Kelly, Toward the molecular mechanism(s) by
which EGCG treatment remodels mature amyloid fibrils, J. Am. Chem. Soc. 135
(2013) 7503–7510.
[72] V.V. Shvadchak, M.M. Claessens, V. Subramaniam, Fibril breaking accelerates alpha-
synuclein fibrillization, J. Phys. Chem. B 119 (2015) 1912–1918.
[73] N. Lorenzen, S.B. Nielsen, Y. Yoshimura, B.S. Vad, C.B. Andersen, C. Betzer, J.D.
Kaspersen, G. Christiansen, J.S. Pedersen, P.H. Jensen, F.A. Mulder, D.E. Otzen, How
epigallocatechin gallate can inhibit alpha-synuclein oligomer toxicity in vitro, J.
Biol. Chem. 289 (2014) 21299–21310.
[74] M. Singh, M. Arseneault, T. Sanderson, V. Murthy, C. Ramassamy, Challenges for re-
search on polyphenols from foods in Alzheimer's disease: bioavailability, metabo-
lism, and cellular and molecular mechanisms, J. Agric. Food Chem. 56 (2008)
4855–4873.