Báo cáo khoa học: Dynamics of a-synuclein aggregation and inhibition of pore-like oligomer development by b-synuclein doc

Dynamics of a-synuclein aggregation and inhibition
of pore-like oligomer development by b-synuclein
Igor F. Tsigelny
1,2
, Pazit Bar-On
3
, Yuriy Sharikov
2
, Leslie Crews
4
, Makoto Hashimoto
3
,
Mark A. Miller
2
, Steve H. Keller
5
, Oleksandr Platoshyn
5
, Jason X J. Yuan
5
and Eliezer Masliah
3,4
1 Departments of Chemistry and Biochemistry, University of California San Diego, La Jolla, CA, USA
2 San Diego Super Computer Center, University of California San Diego, La Jolla, CA, USA
3 Department of Neurosciences, University of California San Diego, La Jolla, CA, USA
4 Department of Pathology, University of California San Diego, La Jolla, CA, USA
5 Department of Medicine, University of California San Diego, La Jolla, CA, USA
In recent years, new hope for understanding the patho-
genesis of Parkinson’s disease (PD) and Lewy body
dementia (LBD) has emerged with the discovery of

mutations and duplications in the a-synuclein (a-syn)
gene that are associated with rare familial forms of
Parkinsonism [1–3]. Moreover, it has been shown that
a-syn is centrally involved in the pathogenesis of both
sporadic and inherited forms of PD and LBD because
this molecule accumulates in Lewy bodies (LBs) [4–6],
synapses, and axons, and its expression in transgenic
(tg) mice [7–9] and Drosophila [10] mimics several
aspects of PD.
The mechanisms through which a-syn leads to neu-
rodegeneration and the characteristic symptoms of
LBD are unclear. However, recent evidence indicates
that abnormal accumulation of misfolded a-syn in the
Keywords
cation channels; modeling; molecular
dynamics; oligomers; synuclein
Correspondence
E. Masliah, Department of Neurosciences,
University of California, San Diego, La Jolla,
CA 92093–0624, USA
Fax: +1 858 5346232
Tel: +1 858 5348992
E-mail:
(Received 1 December 2006, revised 26
January 2007, accepted 8 February 2007)
doi:10.1111/j.1742-4658.2007.05733.x
Accumulation of a-synuclein resulting in the formation of oligomers and
protofibrils has been linked to Parkinson’s disease and Lewy body demen-
tia. In contrast, b-synuclein (b-syn), a close homologue, does not aggregate
and reduces a-synuclein (a-syn)-related pathology. Although considerable

information is available about the conformation of a-syn at the initial and
end stages of fibrillation, less is known about the dynamic process of a-syn
conversion to oligomers and how interactions with antiaggregation chaper-
ones such as b-synuclein might occur. Molecular modeling and molecular
dynamics simulations based on the micelle-derived structure of a-syn
showed that a-syn homodimers can adopt nonpropagating (head-to-tail)
and propagating (head-to-head) conformations. Propagating a-syn dimers
on the membrane incorporate additional a-syn molecules, leading to the
formation of pentamers and hexamers forming a ring-like structure. In con-
trast, b-syn dimers do not propagate and block the aggregation of a-syn
into ring-like oligomers. Under in vitro cell-free conditions, a-syn aggre-
gates formed ring-like structures that were disrupted by b-syn. Similarly,
cells expressing a-syn displayed increased ion current activity consistent
with the formation of Zn
2+
-sensitive nonselective cation channels. These
results support the contention that in Parkinson’s disease and Lewy body
dementia, a-syn oligomers on the membrane might form pore-like struc-
tures, and that the beneficial effects of b-synuclein might be related to its
ability to block the formation of pore-like structures.
Abbreviations
aa, amino acid; a-syn, a-synuclein; b-syn, b-synuclein; GFP, green fluorescent protein; LBD, Lewy body disease; PD, Parkinson’s disease;
POPC, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine; tg, transgenic.
1862 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
synaptic terminals and axons plays an important role
[11–14]. These studies suggest that a-syn oligomers and
protofibrils rather than fibrils might be the neurotoxic
species [15].
a-syn is an abundant presynaptic molecule [16] that
probably plays a role in modulating vesicular synaptic

release [17]. Synucleins belong to a family of related
proteins including a-, b-, and c-synucleins. a-syn
belongs to a class of so-called ‘naturally unfolded pro-
teins’ [13,18]. Such proteins do not have a stable ter-
tiary structure and during their existence change their
conformations. Human a-syn is a 140-amino acid (aa)
protein, and b-syn is a 134-aa protein. Each of the
synucleins is composed of an N-terminal lipid-binding
domain containing 11 residue repeats and a C-terminal
acidic domain that has been proposed to be involved
in protein–protein interactions. It has been shown [19–
22] that at the lipid–protein interface, a-syn has a
conformation characterized by two helical domains
interrupted by a short nonhelical turn. a-syn contains
a highly amyloidogenic hydrophobic domain in the
N-terminus region (aa 60–95), which is partially absent
in b-syn and might explain why b-syn has a reduced
ability to self-aggregate and form oligomers and fibrils
[23,24]. Interestingly, although under physiological
conditions b-syn is nonamyloidogenic, a recent study
demonstrated that certain factors, namely, particular
metals and pesticides, can cause rapid fibrillation of
this molecule and of mixtures of a- and b-syn [25]
under in vitro cell-free conditions. However, previous
studies have shown that in the absence of metals,
b-syn interacts with a-syn and is capable of preventing
a-syn aggregation and related deficits both in vitro and
in vivo [23,24].
Various lines of evidence support the contention that
abnormal aggregates arise from a partially folded

intermediate precursor that contains hydrophobic pat-
ches. It has been proposed that the intermediate a-syn
oligomers form annular protofibrils and pore-like
structures [26–29]. The mechanism through which
monomeric a-syn converts into a toxic oligomer and
later into fibrils is currently under intense investiga-
tion. Recent reviews indicate that the kinetics of a-syn
fibrillation are consistent with a nucleation-dependent
mechanism for which a partially folded intermediate is
needed in the early stages of aggregation [30]. Factors
leading to the formation of the folded intermediates
include oxidation, phosphorylation, mutations, and
lipids in the membrane [30–34]. a-syn oligomerization
might occur on the membrane and involves interac-
tions between hydrophobic residues of the amphipathic
a-helices of a-syn [35]. These studies indicate that the
hydrophobic lipid binding domains in the N-terminal
region might be important in modulating a-syn aggre-
gation [13,36–38]. There are several studies describing
the effects of membranes and membrane-like structure
on aggregation [21,39,40], however, less is known
about the effects of membrane lipids on b-syn struc-
ture. In this context, a recent study has analyzed
by NMR the micelle-bound structure and dynamics of
b- and c-syn [41].
Thus, better understanding of the steps involved in
the process of a-syn aggregation is important in order
to develop intervention strategies that might prevent
or reverse a-syn oligomerization and toxic conversion.
The conformational state of a-syn at the initial and

end stages of fibrillation have been characterized in
some detail and recent studies have shown that early
stage oligomers are globular structures with variable
height (2–6 nm) that after prolonged incubation results
in the formation of elongated protofibrils which disap-
pear upon fibril formation [42].
However, less is known about the dynamic process
of conversion of a-syn at earlier stages and how inter-
actions with antiaggregation chaperones such as b-syn
and heat-shock proteins might occur. This is in part
due to the transient nature of the oligomers and the
difficulties in crystallizing such conformers. Therefore,
the use of computer-based molecular modeling tech-
niques in combination with biochemical and cell based
assays might facilitate understanding the dynamic
characteristics and structure of the synuclein aggre-
gates. In this context, the main objective was to
develop a dynamical model for the early steps of a-syn
aggregation using computer simulations that includes
the process of membrane docking and the potential
mechanisms through which b-syn blocks a-syn aggre-
gation.
Our studies suggest that at early stages, propaga-
ting a-syn dimers immersed in the membrane lead to
the formation of pentamers and hexamers with a
pore-like structure. These ring-like aggregates might
correspond to Zn
2+
-sensitive nonselective cation
channels whose formation is blocked by b-syn. The

inhibitory effect of b-syn may result from its interac-
tion with a-syn, which prevents formation of func-
tional a-syn channels.
Results
Conformational diversity of a-syn and b-syn
molecules during molecular dynamics simulations
To better understand the conformational changes that
a- and b -syn undergo over time and to model the
homo and heterodimeric interactions preventing or
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1863
leading to aggregation, molecular dynamics simula-
tions in water were performed based on the micelle-
bound structure of a-syn as resolved by NMR. This
approach allows the investigation of the dynamic
structural changes of the folded a-syn (micelle-derived)
under simplified conditions. The curved N-terminal
domain of this structure is divided into two regions
(termed helix-N and helix-C) [21] connected by a short
linker (Fig. 1A,B). In our baseline models, the two
curved helical N-terminal domains of the micelle-
derived a-syn molecular structure form an angle
around 55 ± 3° that decreases to around 42–44° dur-
ing the first 2.0 ns of the simulation, and then increa-
ses to 64–70° after 3.0–5.0 ns of simulation. During
simulation (Fig. 1A,B), the initial two curved helical
N-terminal domains (helices N and C) of a-syn trans-
form into three uncurved N-terminal helical structures.
The third helical region appears when the second
curved helix (aa 46–84) converts into two uncurved

helices, helix 2 (aa 46–63) and helix 3 (aa 74–84),
linked by aa 64–73 (Fig. 1A,B). To confirm these
results, we repeated the simulation in water, with dif-
ferent seed numbers, for 3.0 ns. These additional data
corroborate the initial results, showing that over time
a-syn acquires a more three-dimensional shape due to
movement of the C-terminal domain relative to the
N-terminus (Fig. 1A,B). It is worth noting that the
micelle-derived helical structure of a-syn is highly sta-
ble and did not return to an unfolded state even
though the molecular dynamics simulations were per-
formed using the water box to simplify the procedure.
At time zero, b-syn has a structural organization
close to the initial structure of a-syn (Fig. 1C,D).
Unlike a-syn, the curved helix at residues 46–84 of b-
syn does not undergo conversion into two distinct
helices during the course of simulation of up to
5.0 ns. Instead, the curved helices adopt a relatively
straight configuration after 2.0 ns (Fig. 1C,D) and
this conformer increases in stability from 2.0 to
3.5 ns. Further simulation shows additional conform-
ational changes, mostly in the C-terminal tail and the
angle between the N-terminal helices (Fig, 1C,D).
Compared with a-syn, the C-terminal tail of b-syn
displayed greater motility.
Fig. 1. Molecular dynamics simulations of a- and b-syn monomers in water. (A) Snapshots of molecular dynamics conformations of a-syn.
(B) Superimposed a-syn conformers (area of superimposition: aa 1–15). (C) Snapshots of molecular dynamics conformations of b-syn.
(D) Superimposed b-syn conformers (area of superimposition: aa 1–15).
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1864 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS

Further analysis consisted of determining changes in
secondary structure of a-syn and b-syn over time.
After 500 ps of simulation for a-syn a coiled region
appeared, interrupting the a-helix around residue 68
(Fig. S1A). Beginning at 750 ps, turns appeared in the
a-helical structure around residue 47, then after 1.0 ns
this region was transformed into a p-helix (Fig. S1A).
The length of this p-helix increased with time, and
from 3.0 ns covered the region from residues 45–55. In
another part of the sequence, a second p-helix
appeared from 2.0 ns that includes residues 74–83
(Fig. S1A).
Changes in b-syn secondary structure over time con-
sisted of transformations from a bended a-helical
structure to the structure with two straight helices with
further conversion to p-helical structure around residue
30 and the N-terminus region (Fig. S1B). The C-ter-
minal region beyond residue 70 showed limited chan-
ges in secondary structure (Fig. S1B). Overall, b-syn
underwent significantly fewer changes in secondary
structure than a-syn during molecular dynamics simu-
lations.
Interactions of a-syn propagating dimers predict
the formation of pore-like structures
The first studies of the interactions of a-syn were per-
formed by docking the initial structures of two a-syn
monomers on a flat surface without specific limita-
tions. Under these conditions, some low energy com-
plexes of two molecules formed a ‘head-to-tail’
position. This configuration is not favorable for further

aggregation on the membrane. The appearance of such
dimeric aggregates is caused mostly by electric charge
profile complementarities between the N- and C-ter-
mini of a-syn monomers (Fig. 2A and 3A). These
a-syn homodimers can interact with additional a-syn
molecules, but further simulations indicate that the
resulting higher order aggregates are not likely to pro-
duce continuously propagating multimers on the mem-
brane. For the nonpropagating a-syn homodimers,
usually only one a-syn has the membrane binding sur-
face, such as for the 1.5 ns molecular dynamics con-
formers (Fig. 2A and 3A).
Fig. 2. Molecular modeling of nonpropagating and propagating a-syn aggregates on the membrane. (A) a-syn minimal energy nonpropagating
dimers (head-to-tail). (B) a-syn conformer at 4.0 ns oriented to the membrane surface (membrane-contacting residues depicted in orange).
(C–E) Propagating a-syn multimers on the membrane at 3.5 ns (C) Dimer (D) tetramer, and (E) hexamer. Multimers can be formed by dock-
ing of a-syn monomers to a-syn propagating dimers, or by addition of a-syn dimers to a-syn propagating dimers, with either scenario result-
ing in the same final hexamer structure. (F) Final configuration of the hexamer after 3.5 ns on the membrane (side view). (G) Modeling of
multimers at various time points between 1.5 and 4.5 ns (top view). The table to the right margin indicates the inner diameters (ID) and
outer diameters (OD) of the multimers created from the conformers obtained at the various molecular dynamics (MD) time points.
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1865
As previous studies have suggested that the assembly
of a-syn into toxic oligomers might involve interactions
with the membrane [26,43], we proceeded to simulate
the docking of a-syn conformers on a flat surface repre-
senting the membrane. The a-syn conformations at
250 ps increments of molecular dynamics were docked
with their surfaces facing the membrane (defined as
membrane-contacting by the mapas program [44]).
These membrane-contacting surfaces were distributed

as expected along the N-terminal helices of the a-syn
conformers (Fig. 2B). To further verify the conforma-
tional changes of a-syn dimers upon interactions with
the membrane, we conducted docking of two a-syn
4 ps conformers onto a 1-palmitoyl-2-oleoyl-sn-glycero-
3-phosphocholine (POPC) membrane with a grid cell of
1A
˚
, including the membrane in calculations (Fig. 3C).
The electrostatic energy of interaction is around 30–50
kcalÆmol
)1
for docking of two a-syn molecules. Only
minimal differences (< 10% in docking energy values)
were detected between molecules docked on the flat
surface and molecules docked on the POPC membrane.
In general, two possible initial docking configurations
for a-syn molecules on the membrane were observed.
In the first one, the dimer is arranged in a head-to-tail
position and additional monomers cannot easily add to
this complex to propagate toward higher order aggre-
gates, as low-energy binding sites do not appear to exist
for consecutive docking (Fig. 2A and 3A). It is possible
for weakly propagating multimers to form over time up
to 4.0 ns (Fig. 3D), however, the binding energies of
the growing complexes (Fig. 3F) are significantly less
favorable than for propagating configurations (Fig. 3E,
Fig. 3. Modeling of docking of nonpropagating and propagating a-syn dimers and multimers on the membrane. Membrane-contacting N-ter-
minal (n-term) regions are designated by boxes and C-terminal (c-term) regions by lines, as viewed perpendicular to the membrane surface.
For docking, the second a-syn molecule (a-syn 2) docks to the first (a-syn 1), followed by docking of the third a-syn molecule (a-syn 3) to

the second, etc., considering minimal docking energies from all possible docking positions. (A) Non-propagating conformation (head-to-tail) of
two a-syn monomers that prevents low-energy docking of additional monomers. (B) Propagating conformation that allows low energy dock-
ing of additional monomers added sequentially (in the direction of the arrow). (C) Docking of two a-syn conformers at 4.5 ps on the POPC
membrane. (D) Weakly propagating a-syn multimer, composed of four head-to-tail conformers at 4.5 ns (E) Propagating a-syn multimer, com-
posed of five head-to-head conformers at 4.5 ns (F) Electrostatic energies of the complexes growing from one a-syn monomer to a five-
monomer complex at 4.5 ns. The propagating multimer has more favorable electrostatic energy than the weakly propagating multimer.
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1866 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
Fig. S2), and these species would represent only a small
fraction of the total multimers present. Therefore, for
our purposes, this head-to-tail configuration of two
a-syn monomers is designated a ‘nonpropagating
dimer’. In all cases, the nonpropagating interactions
involve regions of the N-terminus up to residue 75 of
one molecule with residues located on the C-terminal
region of the second molecule. In the second configur-
ation, the pair of monomers is oriented ‘head-to-head’
(with tails oriented in similar directions) allowing con-
secutive docking with similar low energy for the
successive molecules of a-syn. We designate this confi-
guration a ‘propagating dimer’ (Figs 2C and 3B,E).
Docking additional a-syn monomers (at specified time
points) to a single initial propagating dimer resulted in
the formation of energetically favorable trimers, tetra-
mers, pentamers, and hexamers on the membrane
(Figs 2C–G, 3E, and Fig. S2). We used the molecular
dynamics conformations ranging from 1.5 to 5.0 ns for
docking, and noted that with longer molecular dynam-
ics simulation times (4.0 ns and later), more residues on
the C-terminal tail (residue 110 and above) became

involved in intermolecular interactions (Fig. S3).
Because the tail of a-syn carries the majority of this
protein’s positive charge, this might help to explain
why there was a significant enhancement of a-syn dimer
docking energies (and accordingly the stability of the
multimers) after 4.0 ns of simulation (Table 1). More-
over, Fig. S2 shows that the most stable conformation
of a-syn occurs after 3.8 ns of molecular dynamics
simulation time. For b-syn, the most stable conforma-
tions arise between 2.2 and 3.5 ns of simulation
(Fig. S2). The most probable a-syn multimers were
selected based on the conformers with the most favora-
ble energies of intermolecular interaction between two
monomers and the most stable conformers. Six distinct
possible multimers were generated as the result of ‘pro-
pagating docking’ (Fig. 2G). These multimers formed
low energy pentamers and hexamers with different con-
figurations that generated ring-like structures with a
central lumen (Fig. 2G). The most stable multimers of
a-syn were generated with a-syn conformers from
4.0 ns simulation and later. The theoretical pentameric
and hexameric conformations of the a-syn multimers
on the membrane are reminiscent of the pore-like
appearance of cell-free a-syn aggregates that have been
reported by atomic force microscopy (AFM) [26].
a-syn propagating dimers form pore-like
structures that are embedded in the membrane
To further investigate how closely the simulation-
derived model resembles a-syn aggregates generated
in vitro, recombinant a-syn was incubated for various

time periods at 65 °C and the preparations analyzed by
western blot and electron microscopy. At 15 h of incu-
bation, immunoblot analysis showed the appearance of
multiple bands at molecular weight levels consistent
with a-syn dimers, trimers, tetramers, and pentamers
(Figs 4A,B). After 20 h, higher order aggregates consis-
tent with hexamers were also detected (Fig. 4A,B).
Ultrastructurally, after 10 h of incubation, ill-defined
globular elements were noted, and around 15 h, ring-
like structures ranging in diameter between 9 and
15 nm with a central channel of 2–5 nm were found
(Fig. 4C–E), while after 20 h fibrils (9–12 nm in diam-
eter) became more apparent (Fig. 4F). Remarkably, the
ring-like structures that formed after 15 h of incubation
were of similar dimension to the a-syn pentamers and
hexamers generated after 4.0 ns of molecular dynamics
simulation (Fig. 4K) and further simulations showed
that they became embedded in the membrane after rel-
atively short (350 ps) molecular dynamics simulation of
the membrane–protein complex (Fig. 5A). During
extended simulation times, the a-syn pentamer embeds
progressively further into the membrane, reaching 16 A
˚
in the membrane by 800 ps (Fig. 5B–E).
b-syn interrupts the formation of propagating
a-syn dimers
We have previously shown that b-syn is capable of
reducing a-syn accumulation and related deficits [23],
however, the molecular characteristics for the interac-
tions between these two molecules are unclear. For this

purpose, we modeled b- and b-syn, and b- and a-syn
heterodimeric interactions. Firstly, theoretical docking
of various molecular dynamics conformers of a-syn to
conformers of
b-syn was performed. All of the docked
a-syn–b-syn complexes displayed a significant level of
negative electrostatic energy of formation (Table 2). In
these simulations, b-syn was able to bind a-syn, cre-
ating stable nonpropagating heterodimers, similar to
nonpropagating a-syn homodimers (Fig. 6A). Strong
Table 1. Intermolecular interaction energies of propagating
a-syn ⁄ a-syn dimers docked on the flat membrane. MD, molecular
dynamics.
MD time (ns) Electrostatic energy (kcalÆmol
)1
)
1.50 )10.6
2.00 )10.6
2.50 )13.4
3.50 )15.1
4.00 )19.7
4.50 )32.9
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1867
electrostatic interactions contributed to the formation
of these a- and b-syn heterodimers. For example, com-
plexes between a-syn (2.5 ns) and b-syn (2.2 ns) dis-
played a minimum intermolecular electrostatic energy
of )31.6 kcalÆmol
)1

, while the electrostatic energy of
interaction between two a-syn (2.5 ns) conformers that
can aggregate into hexamers on the membrane
was )13.4 kcalÆmol
)1
. Thus the binding energy between
a- and b-syn was significantly lower, and more favora-
ble, than the energy of interaction between two a-syn
molecules located on the membrane-like surfaces. The
net charge for b-syn ()16 e) at pH 7.0 is much lower
than that of a-syn (– 9 e), which might help to explain
why it is less likely for b-syn than for a-syn to form
propagating dimers in the membrane [23,24].
In addition to binding to a-syn monomers, the simu-
lations showed that b-syn interacts with a-syn mono-
mers and propagating dimers (Fig. 6C), which can
theoretically form annular-like structures on the mem-
brane. In fact, b-syn binding to earlier a-syn conform-
ers was stronger than that of the next a-syn molecule
that participates in propagating membrane-facing pair-
wise docking. One b-syn molecule (shown in green in
Fig. 6C) docked to an a-syn dimer on the membrane
has a position that conflicts with the neighboring a-syn
molecules that can flank it from both sides in the poss-
ible multimeric complex. Further analysis of the elec-
trostatic energies of interaction between heterodimers
starting with the 1.5 ns molecular dynamics conformer
showed that in most cases, the energy of interaction
between b-syn and a-syn (Table 2) was significantly
lower than for a-syn homodimers ()10.6 kcalÆmol

)1
,
Table 1). This supports the possibility that b-syn might
be able to interrupt the assembly of propagating a-syn
homodimers at various stages of the oligomerization
process.
b-syn blocks the formation of a-syn ring-like
structures and attenuates ion conductance
alterations
Previous studies have shown that when b- and a-syn
are incubated simultaneously, b-syn reduces a-syn
aggregation over time [23,24,45]. However, it is unclear
Fig. 4. Biochemical and ultrastructural analysis of a-syn aggregation, interactions with b-syn, and modeling of ring-like structures. (A) In vitro
cell-free aggregation of a-syn monomers into dimers, trimers, tetramers, pentamers, and hexamers over time without (left panel) and with
(right panel) the addition of b-syn. (B) Semiquantitative analysis of levels of a-syn multimers over time. (C–F) Electron microscopy analysis of
a-syn aggregation over time into ring-like structures and fibrils. (G–J) Electron microscopy analysis demonstrating reduction in a-syn aggrega-
tion over time in the presence of b-syn. Scale bar ¼ 20 nm. (K) Superimposition of a-syn pentamer (4.5 ns) onto the ring-like structure detec-
ted by electron microscopy. Scale bar ¼ 10 nm.
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1868 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
whether b-syn might decrease a-syn aggregation when
added after the process of a-syn oligomerization has
started. The theoretical model presented in the previ-
ous section predicts that under experimental in vitro
conditions, addition of b-syn might prevent further
aggregation of a-syn (Fig. 4). To investigate this possi-
bility, a-syn was allowed to aggregate and then b-syn
was added for various lengths of time. When b-syn
was added 1 h after a-syn aggregation started, there
was a significant decrease in the subsequent formation

of a-syn multimers at the various time points analyzed
(Fig. 4A,B). Consistent with the immunoblot studies,
ultrastructural analysis showed that b-syn reduced the
formation of globular, ring-like, and fibrillar structures
(Fig. 4G–J).
As previous studies have suggested that the a-syn
ring-like structures might form pores in the membrane
that might be responsible for the neurotoxic effects of
a-syn oligomers [26,29,46,47], we investigated whether
abnormally high levels of ion currents are detected in
cells overexpressing a-syn and if this process might be
attenuated by b-syn. For this purpose, we recorded
and compared whole-cell currents in HEK293T cells
transiently transduced with lentiviral vectors expressing
a-syn, b-syn, or a-syn and b-syn together (Fig. 7).
Immunoblot analysis confirmed that cells expressed
comparable levels of a-syn and b-syn (Fig. 7A). Dou-
ble-labeling verified that in cotransduced cells, green
fluorescent protein (GFP) was also expressed with
either a-syn or b-syn (Fig. 7B). The target cells (dis-
playing green fluorescence) for electrophysiological
measurements were identified by cotransduction with a
lenti-GFP vector (Fig. 7C). Cells expressing a-syn
Fig. 5. Modeling of the embedded a-syn
complex in the membrane over time. (A)
Top view (at the level of the uppermost
membrane-associated atom) of the embed-
ded portion of the a-syn pentamer (350 ps)
on the POPC membrane (white, a -syn pen-
tamer; green, membrane phospholipids).

Note the penetration of the pentamer into
the membrane and the exposed membrane
in the center of the a-syn ring-like structure.
(B–E) The steps of penetration of the a-syn
pentamer into the POPC membrane during
0.8 ns molecular dynamics simulation (B, ini-
tial; C, 0.2 ns; D, 0.5 ns; E, 0.8 ns). The
depth of protein insertion into the mem-
brane was measured between the upper-
most membrane-associated atom and the
atom that is embedded deepest into the
membrane.
Table 2. Intermolecular interaction energies of the b-syn conform-
ers with 1.5 ns molecular dynamics a-syn conformer docked on the
flat membrane. MD, molecular dynamics.
b-syn conformer MD time (ns)
Electrostatic energy
(kcalÆmol
)1
)
Initial )27.4
0.25 )22.0
0.50 )29.8
0.75 )45.9
1.00 )37.8
1.50 )26.7
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1869
showed a significant increase in whole-cell currents eli-
cited by depolarizing the cells from a holding potential

of )50 mV to a series of test potentials ranging from
)80 to +80 mV (Fig. 7D,E). The current density at
+80 mV was 54.8 ± 4.3 in cells transduced with an
empty vector, 181.1 ± 18.1 (P<0.001 vs. vector con-
trol) picoamperes ⁄ picofarads in a-syn-expressing cells,
64.2 ± 5.3 (P ¼ 0.21) picoamperes ⁄ picofarads in cells
expressing b-syn, and 78.1 ± 10.4 (P ¼ 0.07) pA ⁄ pF
in cells transduced with a-syn + b-syn (Fig. 7D,E).
Furthermore, the currents in a-syn transduced cells
were sensitive to Zn
2+
(Fig. 7F). Extracellular applica-
tion of 5 lm Zn
2+
reversibly decreased the currents;
the maximal inhibition took place within 3 min
(Fig. 7F). These data indicate that a-syn forms a Zn
2+
-
sensitive nonselective cation channel and that coexpres-
sion of a-syn with b-syn significantly inhibited the
amplitude of currents of the putative a-syn channels.
Discussion
The present study showed by utilizing molecular mode-
ling and molecular dynamics simulations, in combina-
tion with biochemical and ultrastructural analysis, that
a-syn can arrange into homodimers that can adopt
nonpropagating and propagating conformations. The
evidence predicts that propagating a-syn dimers dock
on the membrane surface and can incorporate addi-

tional a-syn molecules, leading to the formation of
pore-like structures. In contrast, b-syn dimers do not
propagate, and when interacting with a-syn aggregates
block the propagation of a-syn into multimeric struc-
tures. Recent studies have suggested that the transfor-
mation of a-syn into a neurotoxic molecule might
involve the sequential conversion of a-syn monomers
into globular oligomers and then protofibrils [46]. In
contrast, a-syn fibrils, which are present in the LBs [6],
might represent a mechanism for isolating toxic oligo-
mers [15]. Previous studies have investigated the con-
formation of a-syn either at the very initial stages of
aggregation [21] or during the process of fibril forma-
tion [42]. In micelles, a-syn monomers consist of two
curved a-helices connected by a short linker in an anti-
parallel arrangement, followed by a short extended
region and a predominantly unstructured mobile tail
[21,48]. The molecular dynamics studies described
here showed that this structure of a-syn displayed
significant changes in the organization of the N-ter-
Fig. 6. Molecular modeling of the interactions of b-syn with a-syn monomers and dimers. (A) a-syn and b-syn minimal energy nonpropagat-
ing heterodimers. (B) Primary electrostatic interactions in the minimal energy a-syn and b-syn dimer. (C) b-syn minimal energy complex with
the a-syn dimer (4.5 ns simulation for a-syn and 2.2 ns simulation for b-syn). This complex does not support further propagation.
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1870 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
minal helices from 2 to 3 helices over time, which
might lead to more complex membrane interactions.
Computerized analysis predicted that these changes
were accompanied over time by alterations in the secon-
dary structure showing that a p-helical conformation

appears in the N-terminus in addition to the a-helix.
However, confirmation of this structural transforma-
tion awaits NMR analysis. Interestingly, molecular
modeling of the misfolding of the Alzheimer’s disease
amyloid-b protein has shown a rapid transition of
the N-terminal a-helix 1 into a p-helix [49]. Such con-
formational changes, in combination with b-hairpin
structures, might be essential to the aggregation
process [50] and the subsequent formation of pore-like
structures.
Under basal conditions, both nonpropagating and
propagating dimers might exist, with a higher propor-
tion of dimers exhibiting a nonpropagating conforma-
tion. In disorders with a-syn aggregation such as PD it
is possible that an increased proportion of propagating
dimers might be present. The conditions that might
favor an increased ratio of propagating a-syn com-
plexes are unclear, but given the conformational insta-
bility of the proteins implied by both experimental and
modeling results, it may be highly sensitive to local
environmental influences. In support of this, closer
association with the membrane has been suggested to
induce a-syn oligomerization [35]. It has been reported
that small oligomeric forms of a-syn preferentially asso-
ciate with lipids and cell membranes [35], however, the
Fig. 7. Studies of ion conductance in a -syn and b-syn transduced cells utilizing lentiviral vectors. (A) HEK293T cells transduced with lentiviral
vectors encoding a-syn, b-syn, and GFP express comparable protein levels. (B) Double-labeling immunocytochemical analysis of 293T cells
cotransduced with lenti-GFP and lenti-asyn or lenti-bsyn. (C) 293T cells transduced with an empty GFP vector, lenti-asyn, or lenti-bsyn and
cells cotransfected with a-syn and b-syn. The transduced cells are indicated by green fluorescence. Scale bar ¼ 20 lm. (D, E) Representa-
tive currents elicited by depolarizing the cells from a holding potential of )50mv to a series of test potentials ranging from )80 to +80mv,

and corresponding current–voltage relationship (E; means ± SE) in tranduced cells. (F) Representative currents at +80mv (left panel) before
(Cont), during (Zn
2+
) and after (Wash) application of 500 l M Zn
2+
. Time course (right panel) of the change in current density before, during,
and after extracellular application of Zn
2+
. The arrows correspond to the currents shown in the left panel (Cont, a; Zn
2+
, b; and Washout, c).
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1871
conformations of a-syn multimers have been difficult to
study due to the inability to crystallize the oligomeric
form of this protein. Our molecular dynamics studies
support the contention that oligomers of a-syn associ-
ate with membranes and suggest that propagating
dimers might b e thermodynamically stable on membranes
in association with lipids. Moreover, the simulations
and modeling suggest that anchoring of propagating
a-syn dimers to the membrane facilitates the incorpor-
ation of additional a-syn monomers, leading to the for-
mation of trimers, tetramers, pentamers, and hexamers,
the latter oligomers forming ring-like structures.
Recent Raman and AFM studies showed that in vitro
early stage oligomers have a globular structure that
elongates over time to form protofibrils [42,51]. High-
resolution ultrastructural and AFM have suggested that
these aggregates might form pore-like structures that

appear to be common to those produced by other mole-
cules involved in neurodegeneration [52], including
amyloid-b protein [53], British dementia peptide (ABri)
[54], Danish dementia peptide (ADan), serum amyloid
A, and amylin [55]. In such studies, the dimension of
the a-syn disc-like structure was 8–10 nm (outer diam-
eter) with a central pore of 1–2 nm [26]. Similarly, our
theoretical studies show that the outer diameter of the
a-syn multimers is 9–15 nm, with an inner diameter of
2–5 nm. Consistent with the possibility that these a-syn
aggregates might represent active pores, cells trans-
duced with a-syn displayed significant increases in
Zn
2+
-sensitive ion channel activity that might corres-
pond to Zn
2+
-sensitive nonselective cation channels.
The Zn
2+
-sensitivity of the a-syn pore-like structures is
likely to be a result of interactions between Zn
2+
and
cysteine, histidine or arginine residues, as has been pre-
viously shown in the case of Zn
2+
-sensitive Ab-derived
pore-like structures [56,57]. Specifically, the His residue
located at position 50 in the a-syn monomer is a poss-

ible candidate for interaction with Zn
2+
ions because it
is situated near the putative pore region of the a-syn
pentamer. The increased ion channel activity observed
in the present study is in agreement with recent results
showing that human neuronal cells expressing mutant
a-syn have high plasma membrane ion permeability
that was sensitive to calcium chelators [47]. Taken
together, these results support the contention that a-syn
aggregates might form functional ion-permeable chan-
nels that in turn might play a role in the mechanisms of
neurodegeneration in LBD.
Therefore, developing strategies that might prevent
a-syn aggregation and subsequent oligomerization into
pore-like structures, or compounds that might block
such potential ion channels could represent a viable
approach to treating disorders characterized by a-syn
aggregation. As in other neurodegenerative disorders,
such as AD similar pore-like structures are formed
[58,59], it is possible that generic antioligomer antibod-
ies that can recognize these assemblies might be useful
[60,61]. The presence of the oligomers in the membrane
might also facilitate recognition by antibodies that pro-
mote clearance of aggregated a-syn [62]. Chaperone
molecules such as heat sock proteins and b-syn might
also be useful. Remarkably, in support of this possibil-
ity, we found that, via interactions with a-syn mono-
mers and multimers, b-syn is capable of preventing
further oligomerization. Moreover, we found that

b-syn ameliorated the abnormal increase in plasma
membrane ion permeability in cells expressing a-syn.
These findings might help to explain previous in vitro
and in vivo studies showing that b-syn is protective
[63]. Therefore, developing molecules that might mimic
the effects of b-syn based on the molecular structure
observations described here may help in the develop-
ment of therapeutic strategies to reduce a-syn aggrega-
tion in disorders with parkinsonism.
Experimental procedures
Molecular dynamics simulations and modeling
of a-syn and b-syn
a-syn is a natively unfolded molecule [64] in the cytosol
that, once it interacts with lipids in the membrane, adopts a
helical structure [21,39,40]. Modeling and simulations were
based on the previously reported NMR structure of the
micelle-bound a-syn (PDB index 1xq8 [21]). The micelle-
derived structure of a-syn was used as the starting point for
the simulations because it was predicted that this conforma-
tion was most likely to lead to oligomerization. For b-syn,
a homology model was generated using the Homology
module of the insight ii program (Accelrys, San Diego,
CA, USA). The molecule was then minimized for 10 000
iterations of steeped descent with the discover program
(Accelrys). Molecular dynamics simulations of a-syn and
b-syn molecules were conducted using periodic boundary
conditions at constant pressure (1 atm) and temperature
(310 K) with the water box in which the shortest dis-
tance between the protein molecule and the box walls was
30 A

˚
. The simulation system contained 237826 atoms, 41
Na
+
and 32 Cl
–
counter ions. The namd molecular
dynamics program version 2.5 [65] was used with the
CHARMM27 force-field parameters [66] to simulate the
behavior of a-syn and b-syn molecules in water under nor-
mal conditions and the interaction of the POPC membrane
with the a-syn aggregates. The temperature was maintained
at 310 K by means of Langevin dynamics using a collision
frequency of 1 ps
)1
. A fully flexible cell at constant pressure
(1 atm) was employed using the Nose
´
–Hoover Langevin
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1872 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
Piston algorithm [67,68], as used in the namd software
package. Initial coordinates were taken from a previously
equilibrated 500 ps system. The van der Waals’ interactions
were switched smoothly to zero over the region 10 A
˚
and
electrostatic interactions were included via the smooth
particle-mesh Ewald summation [69]. The impulse-based
Verlet-I ⁄ r-RESPA method [70,71] was used for multiple

time-stepping: 4 fs for the long-range electrostatic forces, 2 fs
for short-range nonbonded forces, and 1 fs for bonded forces.
The simulation was done in four steps. Initially the sys-
tem of protein and water molecules was minimized for
10 000 iterations. Then the system was heated in 0.1° incre-
ments and equilibrated for 10 ps; then the molecular
dynamics simulation was conducted. Data for analysis were
taken between 50 ps and 5.0 ns of the simulation.
Theoretical docking of synucleins to the
membrane
Interactions between two a-syn and between a-syn and
b-syn monomers were studied using programs hex 4.5
[72,73] and the Docking module of insight ii (Accelrys)
with a 1 A
˚
grid and cut-off distance of 20 A
˚
. Taking into
account previous studies [35] showing that a-syn aggrega-
tion occurs on the membrane, we also studied docking of
synuclein molecules on a flat surface representing the
plasma membrane. Because the docking configurations of
the proteins were mostly open to solution, we used a dielec-
tric constant of 16 to calculate the electrostatic energy for
a-syn docking.
For predicting membrane-contacting surfaces of proteins
we developed the program mapas, based on calculations of
membrane association scores for proteins and protein
aggregates (I. Tsigelny, personal communication). This pro-
gram was applied to intermediate a-syn conformers and top

scoring predicted membrane attractive surfaces were used
to dock the protein molecules and to subsequently calculate
possible a-syn–a-syn docking configurations on the mem-
brane. Low energy a-syn ⁄ a-syn propagating complexes
were used in simulations of consecutive docking of the next
a-syn molecules. We also simulated the behavior of a-syn
complexes on the POPC membrane. The explicit (all-atom)
membrane models were utilized for simulation.
Immunoblot and electron microscopy studies
In previous studies, we have shown that a-syn can prevent
b-syn aggregation in a cell-free system when both synuc-
leins are incubated together at the same time [23]. In this
study, we wanted to determine whether b-syn reduces a-syn
aggregation after a-syn aggregation has already started.
For this purpose, recombinant a-syn (1 lm; Calbiochem,
San Diego, CA, USA) was incubated at 65 °C for time
periods from 1 to 48 h [31]. Incubation at this temperature
allows the study of a-syn aggregation over short periods of
time [31]. After 1 h of incubation recombinant b-syn
(16 lm, purified as previously described [23,63]) was added
to the mix. Samples were subjected to immunoblot analysis
with the mouse monoclonal antibody against a-syn (LB509,
1 : 1000; Zymed Laboratories, San Francisco, CA, USA) as
previously described [31] and analyzed in the versadoc
imaging system using the quantity one software (Bio-Rad,
Hercules, CA, USA).
To investigate the ultrastructural characteristics of the
synuclein aggregates, 1-lL aliquots of a-syn either alone or
in combination with b-syn prepared under identical condi-
tions as for immunoblotting were pipetted onto formvar

coated grids, followed by 2% uranyl acetate staining. Grids
were analyzed with a Zeiss OM 10 electron microscope as
previously described [31].
Preparation and electrophysiological analysis
of cells expressing a- and b-syn
HEK293T cells were grown on 25-mm coverslips at 50%
confluence and were incubated with lentiviruses expressing
a-syn, b-syn, or GFP (each at 1.0 · 10
7
TU) in 10% fetal
calf serum for 24 h at 37 °C, 5% CO
2
. Lentiviruses were
prepared as previously described [74]. The cells were then
washed with NaCl ⁄ P
i
and incubated in Dulbecco’s modi-
fied Eagle’s medium with 10% fetal calf serum for an addi-
tional 4 days. The efficiency of transduction of lenti-GFP
was >90%. For electrophysiological measurements, whole-
cell currents were recorded with an Axopatch-1D amplifier
and a DigiData 1200 interface (Axon Instruments, Sunny-
vale, CA, USA) using patch-clamp techniques. Patch pip-
ettes (2–3 M W) were fabricated on a Sutter electrode puller
using borosilicate glass tubes and fire polished on a Nari-
shige microforge. Command voltage protocols and data
acquisition were performed using pCLAMP-8 software
(Axon Instruments). All experiments were performed at
room temperature (22–24 °C). The ionic composition of the
external solution was (in mm): NaCl 145, KCl 5, MgCl

2
1,
CaCl
2
2, glucose 10, and Hepes 10 (pH ¼ 7.4). During
patch-clamp recording, tetrodotoxin (0.1 lm) and CdCl
2
(0.1 mm) were added to the external solution to block volt-
age-dependent Na
+
and Ca
2+
channels. The ionic compo-
sition of the pipette solution was (in mm): CsCl 150 and
Hepes 10 (pH ¼ 7.2). The current–voltage (I–V) relation-
ship was determined by a step voltage protocol of 50 ms
duration. The membrane potential was held at )50 mV
and stepped to levels between )80 mV and +80 mV in
20-mV increments.
For verification of synuclein expression after lentivirus
infection, transduced cells were harvested in lysis buffer
and analyzed by immunoblot with antibodies against a-syn
(1 : 1000, Chemicon, Temecula, CA, USA), b-syn (prepared
as previously described [75]) and GFP (1 : 1000, Chem-
icon). For immunocytochemistry, cells were cultured on
coverslips until 50% confluence and treated as described
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1873
above, fixed in 4% paraformaldehyde for 20 min, and
blocked overnight at 4 °C in 10% fetal calf serum and 5%

bovine serum albumin.
Acknowledgements
This work was supported by NIH grants AG18440,
HL066012, AG022074, AG5131 and DOE INCITE
grant. The authors are also grateful to IBM for fund-
ing under its Institutes of Innovation program, and for
computational support on its BlueGene computers at
the San Diego Supercomputer Center and at the
Argonne National Laboratory.
References
1 Kruger R, Kuhn W, Muller T, Woitalla D, Graeber M,
Kosel S, Przuntek H, Epplen J, Schols L & Reiss O
(1998) Ala30Pro mutation in the gene encoding a-synuc-
lein in Parkinsons’s disease. Nat Genet 18, 106–108.
2 Polymeropoulos M, Lavedan C, Leroy E, Ide S, Dehejia
A, Dutra A, Pike B, Root H, Rubenstein J, Boyer R
et al. (1997) Mutation in the a-synuclein gene identified
in families with Parkinson’s disease. Science 276,
2045–2047.
3 Singleton AB, Farrer M, Johnson J, Singleton A, Hague
S, Kachergus J, Hulihan M, Peuralinna T, Dutra A,
Nussbaum R et al. (2003) alpha-synuclein locus triplica-
tion causes Parkinson’s disease. Science 302, 841.
4 Takeda A, Mallory M, Sundsmo M, Honer W, Hansen
L & Masliah E (1998) Abnormal accumulation of
NACP ⁄ a-synuclein in neurodegenerative disorders.
Am J Pathol 152, 367–372.
5 Wakabayashi K, Matsumoto K, Takayama K, Yoshi-
moto M & Takahashi H (1997) NACP, a presynaptic
protein, immunoreactivity in Lewy bodies in Parkin-

son’s disease. Neurosci Lett 239, 45–48.
6 Spillantini M, Schmidt M, Lee V-Y, Trojanowski J,
Jakes R & Goedert M (1997) a-synuclein in Lewy bod-
ies. Nature 388, 839–840.
7 Masliah E, Rockenstein E, Veinbergs I, Mallory M,
Hashimoto M, Takeda A, Sagara Sisk A & Mucke L
(2000) Dopaminergic loss and inclusion body formation
in alpha-synuclein mice: implications for neurodegenera-
tive disorders. Science 287, 1265–1269.
8 Lee MK, Stirling W, Xu Y, Xu X, Qui D, Mandir AS,
Dawson TM, Copeland NG, Jenkins NA & Price DL
(2002) Human alpha-synuclein-harboring familial Par-
kinson’s disease-linked Ala-53 fi Thr mutation causes
neurodegenerative disease with alpha-synuclein aggrega-
tion in transgenic mice. Proc Natl Acad Sci USA 99,
8968–8973.
9 Lee VM, Giasson BI & Trojanowski JQ (2004) More
than just two peas in a pod: common amyloidogenic
properties of tau and alpha-synuclein in neurodegenera-
tive diseases. Trends Neurosci 27, 129–134.
10 Feany M & Bender W (2000) A Drosophila model of
Parkinson’s disease. Nature 404, 394–398.
11 Hashimoto M & Masliah E (1999) Alpha-synuclein in
Lewy body disease and Alzheimer’s disease. Brain
Pathol 9, 707–720.
12 Iwatsubo T, Yamaguchi H, Fujimuro M, Yokosawa H,
Ihara Y, Trojanowski JQ & Lee V-M (1996) Purifica-
tion and characterization of Lewy bodies from brains of
patients with diffuse Lewy body disease. Am J Pathol
148, 1517–1529.

13 Lansbury PTJ (1999) Evolution of amyloid: what
normal protein folding may tell us about fibrillogenesis
and disease. Proc Natl Acad Sci USA 96, 3342–3344.
14 Trojanowski J, Goedert M, Iwatsubo T & Lee V (1998)
Fatal attractions: abnormal protein aggregation and
neuron death in Parkinson’s disease and Lewy body
dementia. Cell Death Differ 5, 832–837.
15 Conway KA, Lee SJ, Rochet JC, Ding TT, Williamson
RE & Lansbury PT Jr (2000) Acceleration of oligo-
merization, not fibrillization, is a shared property of
both alpha-synuclein mutations linked to early-onset
Parkinson’s disease: implications for pathogenesis and
therapy. Proc Natl Acad Sci USA 97 , 571–576.
16 Iwai A, Masliah E, Yoshimoto M, De Silva R, Ge N,
Kittel A & Saitoh T (1994) The precursor protein of
non-Ab component of Alzheimer’s disease amyloid
(NACP) is a presynaptic protein of the central nervous
system. Neuron 14, 467–475.
17 Murphy D, Reuter S, Trojanowski J & Lee V-Y (2000)
Synucleins are developmentally expressed, and a-synuc-
lein regulates the size of the presynaptic vesicular pool in
primary hippocampal neurons. J Neurosci 20, 3214–3220.
18 Wright PE & Dyson HJ (1999) Intrinsically unstruc-
tured proteins: re-assessing the protein structure-func-
tion paradigm. J Mol Biol 293, 321–331.
19 Davidson W, Jonas A, Clayton D & George J (1998)
Stabilization of alpha-synuclein secondary structure
upon binding to synthetic membranes. J Biol Chem 273,
9443–9449.
20 Chandra S, Chen X, Rizo J, Jahn R & Sudhof TC

(2003) A broken alpha-helix in folded alpha-synuclein.
J Biol Chem 278, 15313–15318.
21 Ulmer TS, Bax A, Cole NB & Nussbaum RL (2005)
Structure and dynamics of micelle-bound human alpha-
synuclein. J Biol Chem 280, 9595–9603.
22 Kamp F & Beyer K (2006) Binding of alpha-synuclein
affects the lipid packing in bilayers of small vesicles.
J Biol Chem 281, 9251–9259.
23 Hashimoto M, Rockenstein E, Mante M, Mallory M &
Masliah E (2001) b-synuclein inhibits alpha-synuclein
aggregation: a possible role as an anti-parkinsonian
factor. Neuron 32, 213–223.
Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1874 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
24 Uversky VN, Li J, Souillac P, Millett IS, Doniach S,
Jakes R, Goedert M & Fink AL (2002) Biophysical prop-
erties of the synucleins and their propensities to fibrillate:
inhibition of alpha-synuclein assembly by beta- and
gamma-synucleins. J Biol Chem 277, 11970–11978.
25 Yamin G, Munishkina LA, Karymov MA, Lyubchenko
YL, Uversky VN & Fink AL (2005) Forcing nonamy-
loidogenic beta-synuclein to fibrillate. Biochemistry 44,
9096–9107.
26 Ding TT, Lee SJ, Rochet JC & Lansbury PT Jr (2002)
Annular alpha-synuclein protofibrils are produced when
spherical protofibrils are incubated in solution or bound
to brain-derived membranes. Biochemistry 41, 10209–
10217.
27 Rochet JC, Outeiro TF, Conway KA, Ding TT, Volles
MJ, Lashuel HA, Bieganski RM, Lindquist SL & Lans-

bury PT (2004) Interactions among alpha-synuclein,
dopamine, and biomembranes: some clues for under-
standing neurodegeneration in Parkinson’s disease.
J Mol Neurosci 23, 23–34.
28 Lashuel HA, Hartley DM, Petre BM, Wall JS, Simon
MN, Walz T & Lansbury PT Jr (2003) Mixtures of
wild-type and a pathogenic (E22G) form of Abeta40 in
vitro accumulate protofibrils, including amyloid pores.
J Mol Biol 332, 795–808.
29 Volles MJ & Lansbury PT Jr (2002) Vesicle permeabili-
zation by protofibrillar alpha-synuclein is sensitive to
Parkinson’s disease-linked mutations and occurs by a
pore-like mechanism. Biochemistry 41, 4595–4602.
30 Fink AL (2006) The aggregation and fibrillation of
alpha-synuclein. Acc Chem Res 39, 628–634.
31 Hashimoto M, Hernandez-Ruiz S, Hsu L, Sisk A,
Xia Y, Takeda A, Sundsmo M & Masliah E (1998)
Human recombinant NACP ⁄ a-synuclein is aggregated
and fibrillated in vitro: relevance for Lewy body disease.
Brain Res 799, 301–306.
32 Hashimoto M, Takeda A, Hsu LJ, Takenouchi T &
Masliah E (1999) Role of cytochrome c as a stimulator
of a-synuclein aggregation in Lewy body disease. J Biol
Chem 274, 28849–28852.
33 Beyer K (2006) Alpha-synuclein structure, posttransla-
tional modification and alternative splicing as aggrega-
tion enhancers. Acta Neuropathol (Berl) 112, 237–251.
34 Uversky VN (2003) A protein-chameleon: conforma-
tional plasticity of alpha-synuclein, a disordered protein
involved in neurodegenerative disorders. J Biomol Struct

Dyn 21, 211–234.
35 Zhu M, Li J & Fink AL (2003) The association of
alpha-synuclein with membranes affects bilayer struc-
ture, stability, and fibril formation. J Biol Chem 278,
40186–40197.
36 Jao CC, Der-Sarkissian A, Chen J & Langen R (2004)
Structure of membrane-bound alpha-synuclein studied
by site-directed spin labeling. Proc Natl Acad Sci USA
101, 8331–8336.
37 Conway K, Harper J & Lansbury P (1998) Accelerated
in vitro fibril formation by a mutant alpha-synuclein
linked to early-onset Parkinson disease. Nat Med 4,
1318–1320.
38 Uversky VN, Lee HJ, Li J, Fink AL & Lee SJ (2001)
Stabilization of partially folded conformation during
alpha-synuclein oligomerization in both purified and
cytosolic preparations. J Biol Chem 276, 43495–43498.
39 Eliezer D, Kutluay E, Bussell R Jr & Browne G (2001)
Conformational properties of alpha-synuclein in its free
and lipid-associated states. J Mol Biol 307
, 1061–1073.
40 Cole NB, Murphy DD, Grider T, Rueter S, Brasaemle
D & Nussbaum RL (2002) Lipid droplet binding and
oligomerization properties of the Parkinson’s disease
protein alpha-synuclein. J Biol Chem 277, 6344–6352.
41 Sung YH & Eliezer D (2006) Secondary structure and
dynamics of micelle bound beta- and gamma-synuclein.
Protein Sci 15, 1162–1174.
42 Apetri MM, Maiti NC, Zagorski MG, Carey PR &
Anderson VE (2006) Secondary structure of alpha-

synuclein oligomers: characterization by raman and
atomic force microscopy. J Mol Biol 355, 63–71.
43 Volles MJ, Lee SJ, Rochet JC, Shtilerman MD, Ding
TT, Kessler JC & Lansbury PT Jr (2001) Vesicle per-
meabilization by protofibrillar alpha-synuclein: implica-
tions for the pathogenesis and treatment of Parkinson’s
disease. Biochemistry 40, 7812–7819.
44 Reference withdrawn.
45 Lee D, Paik SR & Choi KY (2004) Beta-synuclein
exhibits chaperone activity more efficiently than alpha-
synuclein. FEBS Lett 576, 256–260.
46 Lashuel HA, Petre BM, Wall J, Simon M, Nowak RJ,
Walz T & Lansbury PT Jr (2002) Alpha-synuclein, espe-
cially the Parkinson’s disease-associated mutants, forms
pore-like annular and tubular protofibrils. J Mol Biol
322, 1089–1102.
47 Furukawa K, Matsuzaki-Kobayashi M, Hasegawa T,
Kikuchi A, Sugeno N, Itoyama Y, Wang Y, Yao PJ,
Bushlin I & Takeda A (2006) Plasma membrane ion
permeability induced by mutant alpha-synuclein contri-
butes to the degeneration of neural cells. J Neurochem
97, 1071–1077.
48 Ulmer TS & Bax A (2005) Comparison of structure and
dynamics of micelle-bound human alpha-synuclein and
Parkinson disease variants. J Biol Chem 280,
43179–43187.
49 Urbanc B, Cruz L, Yun S, Buldyrev SV, Bitan G, Tep-
low DB & Stanley HE (2004) In silico study of amyloid
beta-protein folding and oligomerization. Proc Natl
Acad Sci USA 101, 17345–17350.

50 Flock D, Colacino S, Colombo G & Di Nola A (2006)
Misfolding of the amyloid beta-protein: a molecular
dynamics study. Proteins 62, 183–192.
51 Maiti NC, Apetri MM, Zagorski MG, Carey PR &
Anderson VE (2004) Raman spectroscopic characteriza-
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1875
tion of secondary structure in natively unfolded pro-
teins: alpha-synuclein. J Am Chem Soc 126, 2399–2408.
52 Glabe CG (2006) Common mechanisms of amyloid olig-
omer pathogenesis in degenerative disease. Neurobiol
Aging 27, 570–575.
53 Kayed R, Sokolov Y, Edmonds B, McIntire TM,
Milton SC, Hall JE & Glabe CG (2004) Permeabiliza-
tion of lipid bilayers is a common conformation-
dependent activity of soluble amyloid oligomers in
protein misfolding diseases. J Biol Chem 279,
46363–46366.
54 Srinivasan R, Marchant RE & Zagorski MG (2004)
ABri peptide associated with familial British dementia
forms annular and ring-like protofibrillar structures.
Amyloid 11, 10–13.
55 Quist A, Doudevski I, Lin H, Azimova R, Ng D,
Frangione B, Kagan B, Ghiso J & Lal R (2005) Amy-
loid ion channels: a common structural link for pro-
tein-misfolding disease. Proc Natl Acad Sci USA 102,
10427–10432.
56 Kawahara M, Arispe N, Kuroda Y & Rojas E (1997)
Alzheimer’s disease amyloid beta-protein forms Zn(2+)-
sensitive, cation-selective channels across excised mem-

brane patches from hypothalamic neurons. Biophys J
73, 67–75.
57 Bush AI, Multhaup G, Moir RD, Williamson TG,
Small DH, Rumble B, Pollwein P, Beyreuther K &
Masters CL (1993) A novel zinc (II) binding site modu-
lates the function of the beta A4 amyloid protein pre-
cursor of Alzheimer’s disease. J Biol Chem 268,
16109–16112.
58 Glabe CC (2005) Amyloid accumulation and pathogen-
sis of Alzheimer’s disease: significance of monomeric,
oligomeric and fibrillar Abeta. Subcell Biochem 38,
167–177.
59 Arispe N, Pollard HB & Rojas E (1996) Zn
2+
interac-
tion with Alzheimer amyloid beta protein calcium chan-
nels. Proc Natl Acad Sci USA 93, 1710–1715.
60 Kayed R & Glabe CG (2006) Conformation-dependent
anti-amyloid oligomer antibodies. Methods Enzymol
413, 326–344.
61 Kayed R, Head E, Thompson JL, McIntire TM, Milton
SC, Cotman CW & Glabe CG (2003) Common struc-
ture of soluble amyloid oligomers implies common
mechanism of pathogenesis. Science 300, 486–489.
62 Masliah E, Rockenstein E, Adame A, Alford M,
Crews L, Hashimoto M, Seubert P, Lee M, Goldstein J,
Chilcote T et al. (2005) Effects of alpha-synuclein
immunization in a mouse model of Parkinson’s disease.
Neuron 46, 857–868.
63 Hashimoto M, Bar-On P, Ho G, Takenouchi T, Roc-

kenstein E, Crews L & Masliah E (2004) Beta-synuclein
regulates Akt activity in neuronal cells. A possible
mechanism for neuroprotection in Parkinson’s disease.
J Biol Chem 279 , 23622–23629.
64 Weinreb P, Zhen W, Poon A, Conway K & Lansbury
PJ (1996) NACP, a protein implicated in Alzheimer’s
disease and learning, is natively unfolded. Biochem 35,
13709–13715.
65 Kale
´
L, Skeel R, Bhandarkar M, Brunner R, Gursoy A,
Krawetz N, Phillips J, Shinozaki A, Varadarajan K &
Schulten K (1999) NAMD2: greater scalability for
parallel molecular dynamics. J Comp Physics 151,
282–312.
66 Feller S & MacKerell A (2000) An improved empirical
potential energy function for molecular simulations of
phospholipids. J Phys Chem B 104, 7510–7515.
67 Tu K, Tobias DJ & Klein ML (1995) Constant pressure
and temperature molecular dynamics simulation of a
fully hydrated liquid crystal phase dipalmitoylphosphati-
dylcholine bilayer. Biophys J 69, 2558–2562.
68 Feller S, Zhang Y, Pastor RW & Brooks BR (1995)
Constant pressure molecular dynamics simulation: the
Langevin piston method. J Chem Phys 103, 4613–4621.
69 Essmann U, Perera L & Berkowitz ML (1995) A
smooth particle mesh Ewald method. J Chem Phys 103,
8577–8593.
70 Tuckerman M, Berne BJ & Martyna GJ (1992) Reversi-
ble multiple time scale molecular dynamics. J Chem

Phys 97, 1990–2001.
71 Grubmu
¨
ller H, Heller H, Windemuth A & Schulten K
(1991) Generalized Verlet algorithm for efficient molecu-
lar dynamics simulations with long–range interactions.
Mol Simul 6, 121–142.
72 Mustard D & Ritchie DW (2005) Docking essential
dynamics eigenstructures. Proteins: Struc, Func, Bioin-
form 60, 269–274.
73 Ritchie DW & Kemp GJL (2000) Protein docking using
spherical polar Fourier correlations. Proteins: Struct
Func Genet 39, 178–194.
74 Hashimoto M, Rockenstein E, Mante M, Crews L,
Bar-On P, Gage FH, Marr R & Masliah E (2004) An
antiaggregation gene therapy strategy for Lewy body
disease utilizing beta-synuclein lentivirus in a transgenic
model. Gene Ther 11, 1713–1723.
75 Takenouchi T, Hashimoto M, Hsu L, Mackowski B,
Rockenstein E, Mallory M & Masliah E (2001)
Reduced neuritic outgrowth and cell adhesion in neuro-
nal cells transfected with human a-synuclein. Mol Cell
Neurosci 17, 141–150.
Supplementary material
The following supplementary material is available
online:
Fig. S1. Summary of changes in a-syn and b-syn sec-
ondary structure over time. (A) Secondary structure of
a-syn from time 0–5.0 ns. (B) Secondary structure of
b-syn from time 0–5.0 ns.

Modeling of a-syn oligomer formation I. F. Tsigelny et al.
1876 FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS
Fig. S2. Stability of molecular dynamics (MD) simula-
tions of a-syn and b-syn monomers in water. Neigh-
boring MD conformers at 100 ps intervals were
structurally superimposed with the combinatorial
extension program [76] and rmsd between them was
calculated. The most stable MD conformations (boxed
regions of the graph) begin at 3.8 ns for a-syn (red)
and occur between 2.2 and 3.5 ns for b-syn (blue).
Fig. S3. Specific intermolecular interactions between
two head-to-head a-syn monomers over time. Sites of
interaction (arrows) between amino acids in each
of the monomers (a-syn 1 and a-syn 2) at 2.0 ns (A),
3.5 ns (B), 4.0 ns (C), 4.5 ns (D).
This material is available as part of the online article
from
Please note: Blackwell Publishing is not responsible
for the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corres-
ponding author for the article.
I. F. Tsigelny et al. Modeling of a-syn oligomer formation
FEBS Journal 274 (2007) 1862–1877 ª 2007 The Authors Journal compilation ª 2007 FEBS 1877