A causative link between the structure of aberrant protein oligomers and their toxicity

Article (PDF Available)inNature Chemical Biology 6(2):140-7 · February 2010with122 Reads
DOI: 10.1038/nchembio.283 · Source: PubMed
Abstract
The aberrant assembly of peptides and proteins into fibrillar aggregates proceeds through oligomeric intermediates that are thought to be the primary pathogenic species in many protein deposition diseases. We describe two types of oligomers formed by the HypF-N protein that are morphologically and tinctorially similar, as detected with atomic force microscopy and thioflavin T assays, though one is benign when added to cell cultures whereas the other is toxic. Structural investigation at a residue-specific level using site-directed labeling with pyrene indicated differences in the packing of the hydrophobic interactions between adjacent protein molecules in the oligomers. The lower degree of hydrophobic packing was found to correlate with a higher ability to penetrate the cell membrane and cause an influx of Ca(2+) ions. Our findings suggest that structural flexibility and hydrophobic exposure are primary determinants of the ability of oligomeric assemblies to cause cellular dysfunction and its consequences, such as neurodegeneration.
140 NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology
ARTICLE
PUBLISHED ONLINE: 10 JANUARY 2010 | DOI: 10.1038/NCHEMBIO.283
T
he accumulation of specific peptides or proteins as mis-
folded extracellular amyloid fibrils or structurally related
intra cellular inclusions is the hallmark of over 40 human
patho logies, ranging from neurodegenerative disorders (such as,
Alzheimer’s, Parkinsons and prion diseases) to non-neuropathic
systemic amyloidoses (for example, dialysis-related amyloidosis
and light chain amyloidosis) and localized amyloidoses (such as
type II diabetes and atrial amyloidosis)
1
. It is increasingly evident
that, at least in some protein deposition diseases, the pathogenic
species are the oligomeric assemblies that precede the formation
of mature amyloid fibrils
2–8
. A structural characterization of the
oligo mers directed to an understanding of the relationship between
their structural characteristics and their toxic effects is difficult to
achieve, mainly because these species are typically transient and
structurally heterogeneous, thus considerably hampering their
investigation. As a consequence, the structural determinants of the
protein oligomers that are responsible for cell dysfunction are not
yet clear.
The small, 91-residue N-terminal domain of Escherichia coli
HypF (HypF-N) is a stably folded α/β protein with a ferredoxin-
like fold
9
. HypF-N is a valuable model system for investigating
the structural basis of the cellular dysfunction caused by misfolded
protein oligomers. First, monomeric HypF-N is readily able to form
spherical oligomers, protofibrils and amyloid-like fibrils in vitro,
under conditions that destabilize its native state or promote its co-
operative unfolding into partially structured species
10–12
. Second, the
oligomers that form in the early stages of the aggregation process
have the same morphological, structural and tinctorial features as
those formed by disease-related peptides and proteins
11,12
, and they
impair cell viability when added to the extracellular medium of cul-
tured cells
13,14
and when injected into rat brains
15
. Finally, and most
importantly in the context of the present study, HypF-N can rap-
idly be converted into oligomers that remain populated persistently,
rather than transiently, and are sufficiently stable to maintain their
structure and properties even when transferred to conditions that
are very different from those that promote their formation.
In this paper we report the finding that two types of stable oligo-
mers can be formed in vitro from HypF-N that are apparently indis-
tinguishable on the basis of thioflavin T (ThT) binding and display
very similar morphologies when inspected by atomic force micros-
copy (AFM), but that differ in that one induces cellular dysfunction
whereas the other is benign. This observation not only demon-
strates that aggregates with similar morphology can have different
biological properties, but also provides a unique opportunity to
investigate the structural determinants of oligomer toxicity through
comparative studies of their properties. By using protein engineer-
ing and structural mapping by means of N-(1-pyrene)maleimide
(PM) labeling and detection of excimer formation, we have identi-
fied the specific regions of the sequence that are structured in the
toxic and nontoxic oligomers, determined the major differences
between them, and addressed the relationship between oligomer
structure and cellular impairment.
RESULTS
Morphologically similar oligomers show different toxicity
In order to obtain two types of stable oligomeric species, we incu-
bated HypF-N for 4 h at 48 µM, 25 °C in (i) 50 mM acetate buffer,
12% (v/v) trifluoroethanol (TFE), 2 mM DTT, pH 5.5 (condition A)
and (ii) 20 mM trifluoroacetic acid (TFA), 330 mM NaCl, pH 1.7
(condition B). The oligomers were then resuspended in a physio-
logical medium at pH 7.0 with no added co-solvent. Tapping-mode
AFM (TM-AFM) images revealed the presence of spherical bead-
like aggregates with heights in the range of 2–6 nm and 2–7 nm
under conditions A (Fig. 1a) and B (Fig. 1b), respectively. ThT
binding assays showed that both types of oligomers bind this
amyloid-specific dye and increase its fluorescence to similar
extents, indicating the presence of extensive intermolecular β-sheet
structure (Supplementary Results and Supplementary Fig. 1).
1
Department of Biochemical Sciences, University of Florence, Florence, Italy.
2
Department of Physics, University of Genoa, Genoa, Italy.
3
Department
of Chemistry, University of Cambridge, Cambridge, UK.
4
Present addresses: Department of Chemistry and Applied Biosciences, Laboratory of Physical
Chemistry, Eidgenössische Technische Hochschule Zürich, Zurich, Switzerland (S.C.), and Department of Molecular Biology and Biochemistry, University
of California, Irvine, California, USA (A.P.). *e-mail: fabrizio.chiti@unifi.it
A causative link between the structure of aberrant
protein oligomers and their toxicity
Silvia Campioni
1,4
, Benedetta Mannini
1
, Mariagioia Zampagni
1
, Anna Pensalfini
1,4
, Claudia Parrini
2
,
Elisa Evangelisti
1
, Annalisa Relini
2
, Massimo Stefani
1
, Christopher M Dobson
3
, Cristina Cecchi
1
&
Fabrizio Chiti
1
*
The aberrant assembly of peptides and proteins into fibrillar aggregates proceeds through oligomeric intermediates that are
thought to be the primary pathogenic species in many protein deposition diseases. We describe two types of oligomers formed
by the HypF-N protein that are morphologically and tinctorially similar, as detected with atomic force microscopy and thioflavin
T assays, though one is benign when added to cell cultures whereas the other is toxic. Structural investigation at a residue-
specific level using site-directed labeling with pyrene indicated differences in the packing of the hydrophobic interactions
between adjacent protein molecules in the oligomers. The lower degree of hydrophobic packing was found to correlate with a
higher ability to penetrate the cell membrane and cause an influx of Ca
2+
ions. Our findings suggest that structural flexibility
and hydrophobic exposure are primary determinants of the ability of oligomeric assemblies to cause cellular dysfunction and
its consequences, such as neurodegeneration.
© 2010 Nature America, Inc. All rights reserved.
NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology 141
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
These species did not resolubilize when placed under physiological
conditions, as indicated by the preservation of their ability to bind
ThT (Supplementary Fig. 1). Moreover, neither type of oligomer
underwent any detectable structural reorganization following such
a change of the medium (see below).
To assess the biological activity of the spherical aggregates
formed under conditions A and B, we transferred the aggregates to
physiological conditions and added them to the cell culture media
of human neuroblastoma cells (SH-SY5Y) and mouse endothelial
cells (Hend). The state of the cells was first monitored, in each case,
by performing the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-
zolium bromide (MTT) reduction inhibition assay—a generic bio-
chemical test used to monitor cell viability
16
. The aggregates formed
under condition A were found to decrease MTT reduction substan-
tially, relative to untreated cells, in both cell lines (Fig. 1c). Indeed,
their effect is comparable to that of the oligomers formed by Aβ
42
and IAPP (associated with Alzheimer’s disease and type II diabetes,
respectively), which were used here as positive controls
17,18
. By con-
trast, the aggregates formed under condition B had essentially no
effect on the ability of either SH-SY5Y or Hend cells to reduce MTT
(Fig. 1c). No inhibition of MTT reduction was observed when the
cells were treated with native HypF-N (Fig. 1c).
We also stained SH-SY5Y cells with the apoptotic marker Hoechst
33342, which binds to the highly condensed chromatin present in
the nuclei of apoptotic cells, giving rise to a strong fluorescence
signal and allowing the visualization of abnormal nuclei
18,19
.
Fluorescence microscopy images indicate that the nuclei of the cells
treated with HypF-N aggregates formed under condition A were
prominently stained with this dye and often appeared abnormal in
shape (
Fig. 1d), to an extent similar to that of the cells treated with
Aβ
42
and IAPP aggregates (Fig. 1d). By contrast, the nuclei of cells
treated with the aggregates formed under condition B or with the
native protein exhibited a much lower degree of staining, similar
to that of untreated cells (Fig. 1d). Control experiments indicated
that the lack of toxicity of the species formed under condition B is
not due to an insufficient quantity of oligomeric aggregates present
in the experiments (Supplementary Fig. 2). Taken together, these
results indicate that the same polypeptide sequence can assemble
into two distinct types of stable oligomers having similar morpho-
logical and tinctorial properties, but different abilities to cause
cellular dysfunction.
Structural core of the nontoxic HypF-N oligomers
To gain insight into the structural differences of the two types of
oligomers, we expressed 18 mutational variants of the protein, each
carrying a single cysteine residue but located at different positions
along the polypeptide chain, and labeled them with PM. Each
labeled variant was then allowed to aggregate separately under the
different conditions, and the fluorescence spectra of the resulting
samples were acquired and analyzed. PM is a probe of the proxim-
ity between pairs of labeled residues in the species under investiga-
tion, because of its ability to form excited-state dimers, or excimers,
when two PM moieties are within a short distance (about 10 Å)
of each other
20–22
. Excimer formation gives rise to a band in the
430–470 nm region of the PM fluorescence emission spectrum
20–22
;
the presence or absence of this band therefore allows us to deter-
mine whether a labeled position is close (10 Å) or distant (>10 Å),
respectively, to the same labeled position in an adjacent protein
molecule in the oligomer
21,22
. One caveat is that the PM moieties
have the potential to perturb the aggregation process; this potential
problem is, however, avoided in our comparative experiments, as
discussed below.
The fluorescence emission spectra of samples containing
HypF-N labeled with PM at residues 5, 10, 18 and 47 and incubated
120
a
b
c
d
MTT reduction
(% versus untreated cells)
100
80
***
***
**
60
40
20
1234 56
Untreated
cells
HypF-N
condition A
HypF-N
condition B
HypF-N
native
IAPP
[Protein] = 12 µM
Aβ
42
0
Figure 1 | Morphology and toxicity of HypF-N aggregates. (a,b) TM-AFM images (left, height data; right, amplitude data) of HypF-N samples
pre-incubated under conditions A (a) and B (b) and then resuspended at pH 7.0. Scan size, 500 nm. The color bar corresponds to a Z range of 7 nm.
(c
) MTT reduction assay on SH-SY5Y (gray) and Hend (black) cells untreated (lane 1) or treated with Aβ
42
oligomers (lane 2), IAPP oligomers (lane 3),
HypF-N aggregates formed under condition A (lane 4) and B (lane 5), and native HypF-N (lane 6). Error bars correspond to the s.d. values of four
independent experiments. Double and triple asterisks refer to P values lower than 0.01 and 0.001, respectively. (d) Hoechst staining of SH-SY5Y cells
untreated (image 1) or treated with Aβ
42
oligomers (image 2), IAPP oligomers (image 3), HypF-N aggregates formed under condition A (image 4) and
B (image 5) and native HypF-N (image 6). The protein concentration reported refers to monomer concentrations. All images have been acquired at the
same magnification. The scale bar represents 20 µm.
© 2010 Nature America, Inc. All rights reserved.
142 NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
under condition B did not feature an excimer component in the
430–470 nm region (Fig. 2af). In all other cases an excimer peak
centered at about 440 nm was detected in the fluorescence spec-
trum; this peak was, however, of low intensity for protein variants
labeled at residues 22, 28, 34, 40, 65, 69, 75 and 89, but of higher
intensity when residues 25, 55, 59, 79, 83 and 87 were labeled
(
Fig. 2af). These results indicate that for the first group of variants,
those exhibiting no excimer band, the distance between the labeled
residues in adjacent molecules is greater than 10 Å. For the second
group of variants, with weak excimer bands, the labeled residues
are at intermediate distances from each other in the oligomers, or
alternatively in close proximity in only a fraction of the oligomers
that are present in solution. Finally, for the third group of variants,
those with strong excimer bands, the labeled residues are in close
proximity ( 10 Å) to the same residue on one or more adjacent
molecules and indeed are likely to adopt well-defined structure.
From such data we were able to calculate the ratio of the excimer-
to-monomer fluorescence intensities (FI
440nm
/FI
375nm
) for each
labeled residue (Fig. 2g). The profile of excimer ratio versus residue
number suggests that the regions spanning approximately residues
22–34, 55–59 and 75–87 form the structural core of the HypF-N
oligo mers formed under condition B, as the residues in these regions
of the sequence have a high tendency to be in close proximity with
the corresponding residues of adjacent molecules within the
oligomers. The excimer ratio data do not rule out the possibility
that the residues within these regions are also in close proximity
with residues located in other portions of the protein chain (see
below). Nevertheless, they clearly indicate that specific regions of
the molecule are tightly packed within the structural core of these
HypF-N oligomers. Notably, there is excellent agreement between
the locations of these three regions (Fig. 2g) and the three major
peaks in the hydropathy profile of HypF-N (Fig. 2h); our data
indicate, therefore, that the regions of the sequence with the highest
hydrophobicity contribute to the structural core of the nontoxic
aggregates formed by HypF-N.
The core of the toxic oligomers is less tightly packed
The same approach was then used to gain insight into the structure
of the toxic aggregates formed under condition A (Fig. 3af). For
these species, molecules labeled at residues 5, 10, 34, 40, 47 and 89
had no significant excimer components in the spectra (Fig. 3a,c,f).
By contrast, all the remaining positions (18, 22, 25, 28, 55, 59, 65, 69,
75, 79, 83 and 87) showed faint excimer peaks (Fig. 3af). None of
the labeled variants gave rise to spectra with intense excimer bands
(Fig. 3af), indicating that in no case do adjacent molecules pack
together such that their corresponding residues are persistently in
close proximity. A comparison between the values of the excimer
ratios obtained for the toxic and nontoxic aggregates is reported
in Figure 3g. The major differences involve the three hydrophobic
regions of the sequence that yield high values of excimer ratio in
the nontoxic aggregates and much smaller values in the toxic ones
(
Fig. 3g). Differences between the values of excimer ratio in the other
regions are either insignificant or, at most, subtle (Fig. 3g). This
finding indicates that the degree of ordered intermolecular packing
between corresponding hydrophobic regions of adjacent HypF-N
molecules is less marked in the case of the toxic aggregates.
1.0
PM5
PM10
PM18
PM22
PM25
PM28
0.8
0.6
0.4
0.2
4.0
0
–4.0
02040
Residue number
60 80
–8.0
0.6
0.4
0.2
1.0
0.8
0.6
0.4
0.2
Normalized PM fluorescence intensity
Hydropathy
1.0
0.8
0.6
400 440 480 520 400
Wavelength (nm)
440 480 520
0.4
0.2
0
PM34
PM40
PM47
PM55
PM59
PM65
PM83
PM87
PM89
PM69
PM75
PM79
ab
g
h
cd
ef
Excimer ratio
(FI
440 nm
/FI
375 nm
)
Figure 2 | Structural properties of the oligomers formed under condition B. (af) Fluorescence emission spectra of HypF-N oligomers formed under
condition B by protein chains labeled with PM at different positions. Spectra were acquired at 48 µM protein. The color code reported in each panel helps
associate the spectrum with the position of PM labeling. Spectra have been normalized to the intensity of the peak centered at 375 nm. (g
) Ratio between
the fluorescence intensities measured at 440 nm (excimer peak) and 375 nm (monomer peak) for all the labeled positions along the HypF-N sequence.
Error bars correspond to s.d. of at least two independent experiments per position. (h) Hydropathy profile of HypF-N calculated using the Roseman
hydrophobicity scale
44
. The positions of α-helices (red) and β-strands (blue) in the native structure (Protein Data Bank entry 1GXU) are also indicated
as determined by MOLMOL
45
.
© 2010 Nature America, Inc. All rights reserved.
NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology 143
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
We also recorded the fluorescence emission spectra of 1:1 mixtures
of aggregates labeled with PM at different positions (25, 55, 59,
83, 87), prepared under both conditions A and B (Supplementary
Fig. 3a,b). At least in some cases, distinct positions from adjacent
molecules appeared to be spatially close as they gave rise to a medium
or strong excimer signal. However, the excimer ratio values obtained
for the toxic oligomers were considerably lower than those of the
nontoxic oligomers.
To obtain information on the polarity of the environment explored
by the PM moieties in the oligomers formed under conditions A and B,
we calculated the ratio of the fluorescence intensities at 375 nm
(I
I
) and 385 nm (I
III
) for all the labeled positions along the HypF-N
sequence
23
. A value of 2.78 was obtained for the I
I
/I
III
ratio for a
fully soluble and unstructured PM-labeled glutathione dissolved in
water (Supplementary Fig. 4). This value is in agreement with that
reported elsewhere for a fully solvent-exposed PM conjugated to
proteins
24–26
and represents an upper bound for our measurements.
The I
I
/I
III
values obtained for the two types of HypF-N aggregates
indicate that the PM moieties are more accessible to the solvent
(higher I
I
/I
III
ratios) in the toxic oligomers formed under condition
A than in the nontoxic oligomers formed under condition B, parti-
cularly at the level of the three major hydrophobic regions (Fig. 3h).
Moreover, in the case of the oligomers formed under condition B,
there is a good agreement between the labeled positions found
to be buried into a nonpolar environment and the regions of the
sequence that form the structural core of the aggregates—that is,
22–34, 55–59 and 75–87 (Fig. 3g,h). Overall, the results of the
excimer and I
I
/I
III
ratio values indicate that in the toxic oligomers,
the hydrophobic regions are structurally more highly disorganized
and solvent-exposed than they are in the nontoxic oligomers.
The structural differences do not arise from artifacts
In order to interpret the data reported here, it is important to address
the possibility that the PM moieties perturb in some manner the aggre-
gation behavior of HypF-N in one or both of the conditions in which
these studies were carried out. TM-AFM images obtained for HypF-N
samples labeled at positions 10 and 59 and left to aggregate under con-
ditions A and B showed that the globular morphology of the aggre-
gates is preserved upon labeling (
Supplementary Fig. 5). Moreover,
both types of HypF-N oligomers labeled at positions 10, 25, 47 and 59
retained the ability to bind ThT (Supplementary Fig. 6a).
In more structural detail, our analysis is comparative and aims
to explore any significant structural differences between the two
types of aggregates; the differences revealed in the values of excimer
1.0
abg
h
i
cd
ef
Excimer ratio
(FI
440 nm
/FI
375 nm
)
I
I
/I
III
ratio
(FI
375 nm
/FI
385 nm
)
0.6
0.4
0.2
2.5
2.0
1.5
1.0
3,000
2,500
1,500
1,000
ANS fluorescence intensity
at 470 nm (a.u.)
500
455
460
465
470
475
λ
max
(nm)
480
0
050 100 150
[ANS] (µM)
200 250 300
2,000
020406080
Residue number
0.8
0.6
0.4
0.2
1.0
0.8
0.6
0.4
0.2
1.0
Normalized PM fluorescence intensity
0.8
0.6
0.4
0.2
0
400 440 480 520
PM79
PM75
PM69
PM47
PM40
PM34
PM18
PM10
PM5
PM89
PM87
PM83
PM65
PM59
PM55
PM28
PM25
Condition B
PM22
Condition A
400
Wavelength (nm)
440 480 520
Figure 3 | Structural properties of the oligomers formed under condition A. (af) Fluorescence emission spectra of HypF-N oligomers formed under
condition A by protein chains labeled with PM at different positions. Spectra were acquired at 48 µM protein. The color code reported in each panel helps
associate the spectrum with the position of PM labeling. (g
) Excimer ratio profiles of HypF-N oligomers formed under conditions A (red) and B (blue).
Error bars correspond to s.d. of at least two independent experiments per position. (h) Ratio between the fluorescence intensities measured at 375 nm
(I
I
) and 385 nm (I
III
) for all the labeled positions along the HypF-N sequence. Error bars correspond to s.d. of at least two independent experiments. Low
and high ratio values correspond to a nonpolar and highly polar environment, respectively
23
. Values of 2.78 and 2.03 were obtained for glutathione labeled
with PM and dissolved in water and 1-octanol, respectively (Supplementary Fig. 4). Our values are higher than those previously reported for free pyrene
23
.
Similar values have also been obtained by other authors
24–26
and are due to the fact that pyrene is conjugated to a polypeptide chain through a maleimide
moiety rather than being free in solution. (i
) ANS binding to HypF-N oligomers formed under condition A (red) and B (blue). The ANS fluorescence
intensity measured at 470 nm (filled circles) and the wavelength of maximum emission (λ
max
, empty circles) are reported as a function of the ANS
concentration. Protein concentration was 43 µM in all cases.
© 2010 Nature America, Inc. All rights reserved.
144 NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
ratios measured at given positions for the two types of aggregates
are unlikely to be the result of an intrinsic ability of the fluoro-
phore to affect the aggregation process, as this interference would
be revealed in both sets of conditions at the same positions. The
observation that the excimer ratio is higher for the species formed
under condition A at only a small number of positions rules out
the possibility that the ability of PM to perturb the oligomeriza-
tion process is condition-dependent. Moreover, the low values of
the excimer ratios at some positions indicate that PM does not
intrinsically drive the formation of intermolecular interactions
between PM moieties from different HypF-N molecules. As a fur-
ther confirmation, fluorescence emission spectra acquired under
both conditions with PM-labeled β-mercaptoethanol indicated
that the PM moiety is intrinsically soluble under the experimental
conditions used in the present study and does not per se drive the
aggregation process (Supplementary Fig. 6b).
In order to check whether the oligomeric species formed in
conditions A and B varied their structure upon transfer into a
physiological buffer at neutral pH, in which the morphology and
toxicity to cultured cells was assessed, samples of HypF-N labeled
at positions 10, 25, 47, 59, 65 and 83 were separately incubated
in both conditions A and B and analyzed before and after dilu-
tion into 20 mM potassium phosphate buffer, pH 7.0. For each
sample, the resulting PM emission spectra were similar before
and after change of solution conditions, ruling out the possibil-
ity of marked structural changes within the preformed aggregates
(Supplementary Fig. 7).
In another control experiment, the effect of protein concentra-
tion on the excimer ratio values was also tested, indicating that
the difference between toxic and nontoxic aggregates persists
at all protein concentrations examined, including the lowest
(
Supplementary Fig. 3c). Finally, oligomers formed in condi-
tion B using 1:1 mixtures of labeled and unlabeled chains did not
exhibit marked excimer ratio values, indicating that low values
of the excimer ratio for the nontoxic oligomers are obtained only
when the probability of two PM moieties being in close proximity
is reduced (Supplementary Fig. 3d).
Toxic oligomers bind ANS more tightly
To have an independent probe of hydrophobic exposure in the
two types of oligomers, we also tested the ability of the two species
to interact with 8-anilinonaphthalene-1-sulfonate (ANS), a dye
that has been shown to bind to solvent-exposed hydrophobic
a
Untreated cells Native HypF-N
30 µm
30 µm
30 µm
HypF-N condition A
HypF-N condition A
Basal Median Apical
HypF-N condition B
HypF-N condition B
b
efg
hij
cd
Figure 4 | Interaction of the aggregates formed under conditions A and B with cells. (ad) Confocal scanning microscopy images of SH-SY5Y cells
untreated (a) or treated with 12 µM native HypF-N (b), 12 µM HypF-N pre-incubated under condition A (c) and 12 µM HypF-N pre-incubated under
condition B (d). (ej) Optical sections taken through the cells after treatment with 12 µM HypF-N pre-incubated under conditions A (eg) and B (hj)
at basal (e,h), median (f,i) and apical (g,j
) focal lengths. In all images, red and green fluorescence indicates cell profiles and HypF-N, respectively.
HypF-N condition A
Untreated cells
30 µm
HypF-N condition B
Native HypF-N
ab
cd
Figure 5 | Dysregulation of cytosolic Ca
2+
by aggregates formed under
conditions A and B. (ad) Confocal scanning microscopy images of
SH-SY5Y cells showing changes in intracellular free Ca
2+
levels, when cells
are untreated (a) or treated with 12 µ
M native HypF-N (b), 12 µM HypF-N
pre-incubated under condition A (c) and 12 µM HypF-N pre-incubated
under condition B (d). In all images, the green fluorescence arises from the
intracellular Fluo3 probe bound to Ca
2+
.
© 2010 Nature America, Inc. All rights reserved.
NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology 145
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
clusters, and thereby to generate a marked increase in its
fluorescence emission intensity and a blue shift of its maximum
emission wavelength
27,28
. At equivalent ANS concentrations,
the fluorescence intensity measured in the presence of the toxic
oligomers was higher than that obtained with nontoxic oligomers,
indicating a higher degree of exposure to the solvent of hydrophobic
clusters in the toxic species (Fig. 3i). Moreover, the wavelength of
maximum emission (λ
max
) was lower for the toxic oligomers than for
the nontoxic species, indicating that the dye interacts with a more
nonpolar environment (Fig. 3i). No significant ANS binding was
detected for native HypF-N (Supplementary Fig. 8).
HypF-N oligomers differ in cell membrane interaction
It has repeatedly been reported that oligomers can interact with and
permeabilize synthetic lipid vesicles
29
and cell membranes
14
. In order
to assess whether or not the structural differences between the two
types of oligomers detected with PM labeling and ANS binding are
associated with differences in their interactions with the cell mem-
brane, HypF-N aggregates formed under conditions A and B were
added to SH-SY5Y cell culture media. The cells were then exten-
sively washed and examined by confocal scanning microscopy and
immunofluorescence using anti-HypF-N antibodies. Both toxic and
nontoxic oligomers were found to bind to the treated cells; by con-
trast, control experiments carried out with the monomeric protein
showed no interaction (Fig. 4ad). Nevertheless, the imaging of dif-
ferent optical sections of the cells exposed to the oligomers, including
basal, apical and intermediate median planes, revealed the presence
of HypF-N aggregates inside the cells (median planes) only when
these were treated with the toxic oligomers formed under condition A
(Fig. 4ej). The presence of only toxic oligomers inside the cells cannot
be attributed to the different size of these species, as AFM images indi-
cate that the two types of oligomers are submicroscopic and comparable
in size (Fig. 1a,b
). In addition, small green dots were detectable in con-
focal microscopy images outside the cells in both cases (Supplementary
Fig. 9). These results indicate that only the toxic aggregates have suffi-
ciently high structural plasticity and hydrophobic surface to penetrate
the cell membrane and even cross it and reach the interior of the cell.
We then performed a confocal microscope analysis of the cyto-
solic Ca
2+
content in SH-SY5Y cells upon exposure to HypF-N
aggregates. Addition of the toxic oligomers formed under condi-
tion A to the cell culture media was found to cause a large influx
of extracellular Ca
2+
ions into the cytosol, much like the oligomers
formed by other proteins and peptides
30
; by contrast, addition of the
nontoxic oligomers formed under condition B did not cause any
increase of intracellular Ca
2+
ions, as observed upon exposure to the
native protein (Fig. 5).
The ability of both types of oligomers to bind to cell membranes
is consistent with the observation that both can bind to ANS, albeit
with different efficiencies. However, the finding that only the toxic
oligomers cross the hydrophobic bilayer of the cell membrane is
consistent with the conclusions of the PM-labeling and ANS bind-
ing studies that the hydrophobic regions of the toxic oligomers are
more structurally disorganized and solvent exposed than those of
the nontoxic species.
DISCUSSION
It is well known that incubation of the same peptide or protein under
different experimental conditions causes the formation of oligomers
or fibrils with different morphologies and that such differences
result in different degrees of toxicity
31–35
. Similarly, mutations or
covalent modifications can result in different levels of oligomers or
different fibrillar structures with completely different toxicities
36,37
.
However, little experimental information is available on the struc-
tural features of oligomers grown under different conditions and on
the relationship between their structure and their ability to cause
cell dysfunction. In the present study, oligomers formed from the
same protein under different conditions were found to exhibit
similar morphological and tinctorial properties, yet they differ
in their molecular structure and ability to cause cell dysfunction.
Comparisons of the two types of aggregates described here indi-
cate that their structural differences result from different degrees
of packing of the hydrophobic residues within their cores, with
toxicity associated with the level of structural flexibility and solvent-
exposure of such residues. Furthermore, our results indicate that the
toxicity is associated with the ability of the oligomeric species to
form a more pronounced and disruptive interaction with the cells,
stimulating Ca
2+
influx and leading to cell death. Hence, whereas
the ability to form amyloid-like structures is generic to polypeptide
chains, whether or not such species are pathogenic depends on their
structural features, notably the extent to which hydrophobic resi-
dues are flexible and exposed on their surfaces within the environ-
ment of a living organism.
Our ability to characterize these two forms of oligomers can be
attributed to the high aggregation propensity of HypF-N, which
has allowed oligomeric states that would normally be metastable
to be trapped. Our findings, however, do not seem to be limited
to the HypF-N aggregates, and could indeed explain the toxic
properties of the oligomers formed by disease-related systems.
Hydrogen/deuterium (H/D) exchange measurements carried out
on the protofibrils and mature fibrils formed by the 40-residue
form of the amyloid-β peptide (Aβ
40
), the peptide associated with
Alzheimer’s disease, indicate that the degree of protection from H/D
exchange, observed at the level of the two regions of the sequence
forming the structural core of both types of aggregates, is higher
in the fibrils than in the protofibrils
38
. Moreover, in the case of the
fibrils, some of the residues flanking these two regions are also pro-
tected from H/D exchange
38
. Notably, these two regions correspond
to the highest peaks in the hydropathy profile of this peptide, sug-
gesting that the higher degree of structure of these hydrophobic
portions of the sequence in the fibrils could explain why mature
Aβ
40
fibrils are less toxic than protofibrils
6
. Other results also indi-
cate a correlation between the size and surface hydrophobicity of
Aβ
40
aggregates and their ability to decrease the bilayer fluidity of
model membranes
39
, suggesting that the exposure to the solvent
of hydrophobic surfaces determines the ability of these species to
interact with cell membranes. A correlation between hydrophobic-
ity, tendency to form aggregates and aggregate cytotoxicity has also
been observed in comparative studies where the behavior of different
homopolymeric amino acid (HPAA) stretches was investigated
40,41
.
It has been recently reported that expanded huntingtin-exon1
forms fibrillar aggregates at two different temperatures that have
different structural and physical properties as well as different cyto-
toxicities
35
. The structures and toxicities of both forms of the aggre-
gates are comparable with those extracted from regions of mouse
brains affected to different extents by huntingtin deposition. In both
pairs of structures, a direct relationship between structural flexibil-
ity and cytotoxicity of amyloid assemblies was found, supporting
the generality of our conclusions
35
.
In conclusion, the data obtained here lend support to the idea
that a key feature in the generation of toxicity is the conversion of a
species of aggregates in which stability is associated with extreme
burial of hydrophobic residues to one in which such residues are
substantially exposed and disorganized
42
. On the same grounds,
we suggest that for therapeutic purposes the toxicity can be
substantially reduced if the hydrophobic residues are incorpo-
rated to a greater extent within the interior of the oligomeric
assemblies, even in the absence of an effective change in mor-
phology. From a broader perspective, our data establish a link
between bio logical activity of aberrant protein oligomers and
precise structural features within them. They suggest that
solvent-exposed and structurally disorganized hydrophobic
residues within small protein oligomers are at the origin of the
© 2010 Nature America, Inc. All rights reserved.
146 NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
pathogenesis of important debilitating human diseases and can
represent important molecular targets for therapeutic intervention.
METHODS
Protein expression, purification and mutagenesis. Wild-type and mutated
HypF-N production was carried out as described in Supplementary Methods.
Purified proteins were stored at −20 °C in 5 mM acetate buffer, pH 5.5, with 2 mM
dithiothreitol (DTT). Mutations in the gene coding for wild-type HypF-N were
generated using the QuickChange site-directed mutagenesis kit (Stratagene), as
described in Supplementary Methods. The molecular mass of the purified vari-
ants was checked with MALDI-MS. The protein purity was found by SDS-PAGE
to be >95% in all cases. The purified variants were stored at −20 °C in 100 mM
potassium phosphate buffer, pH 7.0, with 2 mM tris(2-carboxyethyl)phosphine
hydrochloride.
Preparation of HypF-N oligomers. Oligomers were prepared by diluting the
protein stock solution to 48 µ
M in (i) 50 mM acetate buffer, 12% (v/v) trifluoro-
ethanol (TFE), 2 mM DTT, pH 5.5 (condition A) and (ii) 20 mM trifluoroacetic
acid (TFA), 330 mM NaCl, pH 1.7 (condition B). The resulting samples were
incubated for 4 h at 25 °C.
TM-AFM. Aggregates were centrifuged at 16,100 r.c.f. for 10 min and resuspended
in potassium phosphate buffer, pH 7.0. 10 µl of each sample was diluted 100 times
(or 500 times for samples not resuspended at pH 7.0), deposited on a freshly
cleaved mica substrate and dried under vacuum. TM-AFM images were acquired
in air using a Dimension 3100 SPM with a ‘G’ scanning head (maximum scan
size 100 µm) and driven by a Nanoscope IIIa controller, and a Multimode SPM
equipped with ‘E’ scanning head (maximum scan size 10 µm) and driven by a
Nanoscope IV controller (Digital Instruments, Veeco). Single beam uncoated silicon
cantilevers (type OMCL-AC160TS, Olympus) were used. The drive frequency
was between 320 and 340 kHz; the scan rate was 0.5–2.0 Hz. Aggregate sizes were
measured from the height in cross section of the topographic AFM images; due to
the drying procedure applied, the measured heights were multiplied by a shrinking
factor of 2.2 and evaluated comparing the heights of native HypF-N under liquid
and after drying.
MTT assay and Hoeschst staining test. Human SH-SY5Y neuroblastoma cells
and mouse Hend endothelium cells were cultured as described in Supplementary
Methods. Both types of HypF-N aggregates were centrifuged at 16,100 r.c.f., dried
under N
2
to remove TFE when necessary, dissolved in DMEM without phenol red
and immediately added to the cell media of SH-SY5Y and Hend cells for 24 h, at 12 µM
monomer concentration. Aggregates of Aβ
42
and IAPP (refs. 17,18) were tested at
the same concentration. HypF-N aggregates formed under condition B were also
tested at 2 and 48 µM. Native HypF-N was tested by diluting the stock solution of
HypF-N to a final protein concentration of 12 µM. Cytotoxicity was assessed by the
MTT assay as reported in Supplementary Methods. SH-SY5Y cells treated
for 24 h with 12 µM of the aggregates formed by HypF-N, Aβ
42
and IAPP or
12 µM native HypF-N were stained with the Hoechst 33342 dye as described in
Supplementary Methods.
Labeling with PM. Each protein variant was diluted to 0.2 mM in 100 mM
potassium phosphate, pH 7.0, 3 M guanidine hydrochloride. Aliquots of PM
(Molecular Probes) in DMSO were added to a tenfold molar excess of dye. The
sample was left in the dark on a shaker for 1 h at 37 °C, then overnight at 4 °C.
The reaction was quenched with 5 µl of trifluoroacetic acid. The unbound dye
was removed by extensive dialysis (3.0 kDa cutoff), and the sample was centrifuged
to remove any precipitate. The concentration of PM was determined using ε
344nm
=
40,000 M
–1
cm
–1
(ref. 43). Protein concentration was measured at 280 nm after
subtraction of the contribution of PM.
Pyrene fluorescence emission spectra. Fluorescence emission spectra of the type A
and B PM-labeled aggregates were measured at 25 °C on a PerkinElmer LS 55
spectrofluorimeter with an excitation of 344 nm. Protein concentration was 48 µM.
The spectra were smoothed and normalized to the intensity of the peak at 375 nm.
ANS titration. Both types of HypF-N aggregates formed at 48 µM were centrifuged
for 10 min at 16,100 r.c.f. The pellets were resuspended in 20 mM potassium phos-
phate buffer, pH 7.0. Aliquots of ANS from a stock solution in 20 mM potassium
phosphate buffer, pH 7.0, were subsequently added to the aggregates, to a final
ANS concentration ranging from 0 to 237 µM. The final protein concentration
was 43 µM in all cases. The spectra were immediately acquired and processed as
described in Supplementary Methods. The difference between the resulting fluo-
rescence intensity at 470 nm and that measured with only protein in the absence of
ANS was used as the effective ANS fluorescence.
Interaction of the aggregates with the cell membrane. SH-SY5Y cells were seeded
on glass coverslips and analyzed using a Leica TCS SP5 confocal scanning micro-
scope, equipped with an argon laser source. Cells were first treated for 60 min at
37 °C with native HypF-N or with aggregates formed under conditions A and B,
after centrifugation at 16,100 r.c.f. and resuspension of the pellet in DMEM without
phenol red. The final protein concentration was 12 µM. The cells were washed with
phosphate-buffered saline, counterstained for 10 min with 50 µg ml
–1
Alexa Fluor
633–conjugated wheat germ agglutinin and fixed in 2% (w/v) buffered
paraformaldehyde for 10 min at room temperature (20 °C). After plasma
membrane permeabilization with a 3% (v/v) glycerol solution for 5 min, the cover-
slips were incubated for 60 min with 1:1,000 diluted rabbit polyclonal anti-HypF-N
antibodies (Primm srl) and then for 90 min with 1:1,000 diluted Alexa Fluor
488–conjugated anti-rabbit secondary antibodies.
Cytosolic Ca
2+
dysregulation. SH-SY5Y cells seeded on glass coverslips were
treated with native HypF-N or aggregates as described above. Cells were then
washed and loaded for 30 min at 37 °C with 10 µM Fluo3-AM, 0.01% (w/v)
pluronic acid F-127 in Hank’s Balanced Salt Solution. Cells were then washed and
fixed in 2.0% (w/v) buffered paraformaldehyde for 10 min at room temperature.
The analysis was performed using the confocal scanning system described above.
Accession codes. Protein Data Bank: The native structure of HypF-N was
deposited as part of a previous study under accession code 1GXU.
Received 24 September 2009; accepted 6 November 2009;
published online 10 January 2010
References
1. Chiti, F. & Dobson, C.M. Protein misfolding, functional amyloid, and human
disease. Annu. Rev. Biochem. 75, 333–366 (2006).
2. Janson, J., Ashley, R.H., Harrison, D., McIntyre, S. & Butler, P.C. e
mechanism of islet amyloid polypeptide toxicity is membrane disruption
by intermediate-sized toxic amyloid particles. Diabetes 48, 491–498
(1999).
3. Sousa, M.M. & Saraiva, M.J. Neurodegeneration in familial amyloid
polyneuropathy: from pathology to molecular signaling. Prog. Neurobiol. 71,
385–400 (2003).
4. Merlini, G. & Bellotti, V. Molecular mechanisms of amyloidosis. N. Engl. J.
Med. 349, 583–596 (2003).
5. Silveira, J.R. et al. e most infectious prion protein particles. Nature 437,
257–261 (2005).
6. Rahimi, F., Shanmugam, A. & Bitan, G. Structure-function relationships of
pre-brillar protein assemblies in Alzheimer’s disease and related disorders.
Curr. Alzheimer Res. 5, 319–341 (2008).
7. Cookson, M.R. & van der Brug, M. Cell systems and the toxic mechanism(s)
of alpha-synuclein. Exp. Neurol. 209, 5–11 (2008).
8. Kayed, R. et al. Common structure of soluble amyloid oligomers implies
common mechanism of pathogenesis. Science 300, 486–489 (2003).
9. Rosano, C. et al. Crystal structure and anion binding in the prokaryotic
hydrogenase maturation factor HypF acylphosphatase-like domain. J. Mol.
Biol. 321, 785–796 (2002).
10. Chiti, F. et al. Solution conditions can promote formation of either amyloid
protolaments or mature brils from the HypF N-terminal domain. Protein
Sci. 10, 2541–2547 (2001).
11. Marcon, G. et al. Amyloid formation from HypF-N under conditions in
which the protein is initially in its native state. J. Mol. Biol. 347, 323–335
(2005).
12. Campioni, S. et al. Conformational properties of the aggregation precursor
state of HypF-N. J. Mol. Biol. 379, 554–567 (2008).
13. Bucciantini, M. et al. Inherent toxicity of aggregates implies a common
mechanism for protein misfolding diseases. Nature 416, 507–511 (2002).
14. Bucciantini, M. et al. Prebrillar amyloid protein aggregates share common
features of cytotoxicity. J. Biol. Chem. 279, 31374–31382 (2004).
15. Baglioni, S. et al. Prebrillar amyloid aggregates could be generic toxins in
higher organisms. J. Neurosci. 26, 8160–8167 (2006).
16. Mosmann, T. Rapid colorimetric assay for cellular growth and survival:
application to proliferation and cytotoxicity assays. J. Immunol. Methods 65,
55–63 (1983).
17. Cecchi, C. et al. Seladin-1/DHCR24 protects neuroblastoma cells against
Abeta toxicity by increasing membrane cholesterol content. J. Cell. Mol. Med.
12, 1990–2002 (2008).
18. Cecchi, C. et al. Replicating neuroblastoma cells in dierent cell cycle phases
display dierent vulnerability to amyloid toxicity. J. Mol. Med. 86, 197–209
(2008).
19. Downs, T.R. & Wilnger, W.W. Fluorometric quantication of DNA in cells
and tissue. Anal. Biochem. 131, 538–547 (1983).
20. Birks, J.B. Excimers and exciplexes. Nature 214, 1187–1190 (1967).
21. Hammarström, P. et al. Structural mapping of an aggregation nucleation site
in a molten globule intermediate. J. Biol. Chem. 274, 32897–32903 (1999).
22. Krishnan, R. & Lindquist, S.L. Structural insights into a yeast prion
illuminate nucleation and strain diversity. Nature 435, 765–772 (2005).
23. Dong, D.C. & Winnik, M.A. e Py scale of solvent polarities. Can. J. Chem.
62, 2560–2565 (1984).
© 2010 Nature America, Inc. All rights reserved.
NATURE CHEMICAL BIOLOGY | VOL 6 | FEBRUARY 2010 | www.nature.com/naturechemicalbiology 147
ARTICLE
NATURE CHEMICAL BIOLOGY DOI: 10.1038/NCHEMBIO.283
24. Christensen, P.A., Pedersen, J.S., Christiansen, G. & Otzen, D.E. Spectroscopic
evidence for the existence of an obligate pre-brillar oligomer during
glucagon brillation. FEBS Lett. 582, 1341–1345 (2008).
25. Hammarström, P. et al. Structural mapping of an aggregation nucleation site
in a molten globule intermediate. J. Biol. Chem. 274, 32897–32903 (1999).
26. irunavukkuarasu, S., Jares-Erijman, E.A. & Jovin, T.M. Multiparametric
uorescence detection of early stages in the amyloid protein aggregation of
pyrene-labeled alpha-synuclein. J. Mol. Biol. 378, 1064–1073 (2008).
27. Semisotnov, G.V. et al. Study of the “molten globuleintermediate state in
protein folding by a hydrophobic uorescent probe. Biopolymers 31, 119–128
(1991).
28. Cardamone, M. & Puri, N.K. Spectrouorimetric assessment of the surface
hydrophobicity of proteins. Biochem. J. 282, 589–593 (1992).
29. Kayed, R. et al. Permeabilization of lipid bilayers is a common conformation-
dependent activity of soluble amyloid oligomers in protein misfolding
diseases. J. Biol. Chem. 279, 46363–46366 (2004).
30. Demuro, A. et al. Calcium dysregulation and membrane disruption as a
ubiquitous neurotoxic mechanism of soluble amyloid oligomers. J. Biol.
Chem. 280, 17294–17300 (2005).
31. Petkova, A.T. et al. Self-propagating, molecular-level polymorphism in
Alzheimer’s beta-amyloid brils. Science 307, 262–265 (2005).
32. Inaba, S., Okada, T., Konakahara, T. & Kodaka, M. Specic binding of
amyloid-beta-protein to IMR-32 neuroblastoma cell membrane. J. Pept. Res.
65, 485–490 (2005).
33. Lee, S., Fernandez, E.J. & Good, T.A. Role of aggregation conditions in
structure, stability, and toxicity of intermediates in the Abeta bril formation
pathway. Protein Sci. 16, 723–732 (2007).
34. Kayed, R. et al. Annular protobrils are a structurally and functionally
distinct type of amyloid oligomer. J. Biol. Chem. 284, 4230–4237 (2009).
35. Nekooki-Machida, Y. et al. Distinct conformations of in vitro and in vivo
amyloids of huntingtin-exon1 show dierent cytotoxicity. Proc. Natl. Acad.
Sci. USA 106, 9679–9684 (2009).
36. Yoshiike, Y., Akagi, T. & Takashima, A. Surface structure of amyloid-beta
brils contributes to cytotoxicity. Biochemistry 46, 9805–9812 (2007).
37. Hung, L.W. et al. Amyloid-beta peptide (Abeta) neurotoxicity is modulated
by the rate of peptide aggregation: Abeta dimers and trimers correlate with
neurotoxicity. J. Neurosci. 28, 11950–11958 (2008).
38. Kheterpal, I. & Wetzel, R. Hydrogen/deuterium exchange mass spectrometry—a
window into amyloid structure. Acc. Chem. Res. 39, 584–593 (2006).
39. Kremer, J.J., Pallitto, M.M., Sklansky, D.J. & Murphy, R.M. Correlation of
β-amyloid aggregate size and hydrophobicity with decreased bilayer uidity
of model membranes. Biochemistry 39, 10309–10318 (2000).
40. Oma, Y., Kino, Y., Sasagawa, N. & Ishiura, S. Intracellular localization of
homopolymeric amino acid-containing proteinexpressed in mammalian cells.
J. Biol. Chem. 279, 21217–21222 (2004).
41. Oma, Y., Kino, Y., Sasagawa, N. & Ishiura, S. Comparative analysis of the
cytotoxicity of homopolymeric amino acids. Biochim. Biophys. Acta 1748,
174–179 (2005).
42. Cheon, M. et al. Structural reorganisation and potential toxicity of
oligomeric species formed during the assembly of amyloid brils. PLoS
Comput. Biol. 3, 1727–1738 (2007).
43. Haugland, R.P. iol-reactive probes excited with ultraviolet light. in
Handbook of Fluorescent Probes and Research Products 9
th
edn. (ed. Gregory, J.)
95 (Molecular Probes, Eugene, Oregon, USA, 2002).
44. Roseman, M.A. Hydrophilicity of polar amino acid side-chains is markedly
reduced by anking peptide bonds. J. Mol. Biol. 200, 513–522 (1988).
45. Koradi, R., Billeter, M. & Wüthrich, K. MOLMOL: a program for display and
analysis of macromolecular structures. J. Mol. Graph. 14, 51–55 (1996).
Acknowledgments
This work was supported by the European Union (Project EURAMY), by the Italian
Ministero dell’Istruzione, Università e Ricerca (FIRB RBNE03PX83, PRIN 2006058958
and PRIN 2007XY59Z), Fondazione Cariplo, Ente Cassa di Risparmio di Firenze (project
“Lipid rafts” 2008) and the European Molecular Biology Organization Young Investigator
Programme. We are grateful to S. Torrassa and D. Nichino for assistance with AFM
measurements and to I. Shalova for assistance in the purification and labeling of some of
the variants used in this work.
Author contributions
S.C. and F.C. designed the variants and the experiments to investigate the structure of
HypF-N oligomers. S.C. (in part also B.M.) produced, purified and labeled all the vari-
ants and performed all the fluorescence experiments (PM, ThT and ANS). A.P. and E.E.
performed the MTT assays. M.Z. performed the confocal microscopy experiments. C.P.
and A.R. designed, performed and analyzed the AFM data. M.S. and C.C. supervised the
experiments on cell cultures. F.C. supervised all the experiments. S.C., C.M.D. and F.C.
wrote the manuscript with contributions from A.R., C.C. and M.S.
Additional information
Supplementary information is available online at http://www.nature.com/
naturechemicalbiology/. Reprints and permissions information is available online at
http://npg.nature.com/reprintsandpermissions/. Correspondence and requests for
materials should be addressed to F.C.
© 2010 Nature America, Inc. All rights reserved.
    • "An interesting exemplification of the intimate interplay between oligomer structure and cell membrane composition is offered by the behaviour of the amyloid-forming protein domain HypF-N. It has been shown that the toxicity of HypF-N oligomers depends on both the oligomer structure and the physicochemical properties of the lipid membrane arising from its lipid composition [14, 16, 17, 19]. Only the oligomeric form here called OA appeared to be toxic to cultured neuronal cells, cultured primary neurons and whole animal models, resembling the effects of toxic Aβ oligomers, while the oligomeric form here called OB was not found to significantly affect cell viability and animal cognitive abilities [14][15][16] 18]. "
    [Show abstract] [Hide abstract] ABSTRACT: Protein misfolded oligomers are considered the most toxic species amongst those formed in the process of amyloid formation and the molecular basis of their toxicity, although not completely understood, is thought to originate from the interaction with the cellular membrane. Here, we sought to highlight the molecular determinants of oligomer-membrane interaction by atomic force microscopy. We monitored the interaction between multiphase supported lipid bilayers and two types of HypF-N oligomers displaying different structural features and cytotoxicities. By our approach we imaged with unprecedented resolution the ordered and disordered lipid phases of the bilayer and different oligomer structures interacting with either phase. We identified the oligomers and lipids responsible for toxicity and, more generally, we established the importance of the membrane lipid component in mediating oligomer toxicity. Our findings support the importance of GM1 ganglioside in mediating the oligomer-bilayer interaction and support a mechanism of oligomer cytotoxicity involving bilayer destabilization by globular oligomers within GM1-rich ordered raft regions rather than by annular oligomers in the surrounding disordered membrane domains.
    Full-text · Article · Jul 2016
    • "Further, mutant SOD1 accumulation in mitochondria is both necessary and sufficient to induce mitochondrial damage in cultured cells [15,16] and ALS-like phenotypes in model mice [17] supporting a tight connection between mitochondrial localization and toxicity of SOD1. Why this enhanced localization is detrimental to mitochondria is less clear, although it is plausible that oligomeric assemblies can readily interact with cell membranes, including mitochondrial membranes [18,19], thus affecting mitochondrial physiology. The interaction of misfolded mutant SOD1 with key factors such as the mitochondrial channel protein VDAC1, and the ability of SOD1 to impact on the overall mechanisms of protein import into mitochondria suggest that the proper exchange of ions and proteins between mitochondria and cytosol might be particularly targeted by mitochondria-associated misfolded SOD1 species [20,21]. "
    [Show abstract] [Hide abstract] ABSTRACT: Alterations in the structure and functions of mitochondria are a typical trait of Amyotrophic Lateral Sclerosis, a neurodegenerative disease characterized by a prominent degeneration of upper and lower motor neurons. The known gene mutations that are responsible for a small fraction of ALS cases point to a complex interplay between different mechanisms in the disease pathogenesis. Here we will briefly overview the genetic and mechanistic evidence that make dysfunction of mitochondria a candidate major player in this process.
    Article · Jul 2016
    • "In addition, Aβ1–42 protofibrils but not fibrils were shown to stimulate microglial production of tumor necrosis factor α suggesting a role for soluble Aβ aggregates in stimulating inflammatory responses and toxicity (Paranjape et al., 2012). Oligomers of different proteins were reported to take a common sequence-independent conformation which suggests that a similar mechanism of toxicity would exist for all the amyloid diseases (Kayed and Glabe, 2006; Chiti and Dobson, 2006; Campioni et al., 2010). Aβ oligomers exert their toxicity through a variety of mechanisms including receptor and direct membrane interactions, reviewed by Kayed and Lasagna-Reeves (2013). "
    [Show abstract] [Hide abstract] ABSTRACT: The incidence of Alzheimer's disease (AD) is growing every day and finding an effective treatment is becoming more vital. Amyloid-β (Aβ) has been the focus of research for several decades. The recent shift in the Aβ cascade hypothesis from all Aβ to small soluble oligomeric intermediates is directing the search for therapeutics towards the toxic mediators of the disease. Targeting the most toxic oligomers may prove to be an effective treatment by preventing their spread. Specific targeting of oligomers has been shown to protect cognition in rodent models. Additionally, the heterogeneity of research on Aβ oligomers may seem contradictory until size and conformation are taken into account. In this review, we will discuss Aβ oligomers and their toxicity in relation to size and conformation as well as their influence on inflammation and the potential of Aβ oligomer immunotherapy.
    Full-text · Article · Apr 2016
Show more

    We use cookies to give you the best possible experience on ResearchGate. Read our cookies policy to learn more.