Fibrillisation of ring-closed amyloid peptides

Article abstract of DOI:10.1039/c2cc17583e

Ring-closing olefin metathesis reactions are used to create intra-molecularly ring closed peptides or inter-molecularly ring-closed peptide dimers based on a designed amyloid peptide sequence. The uncrosslinked peptide self-assembles into high aspect ratio nanotubes, however ring-closing leads to the formation of fibrillar and twisted/helical ribbon structures. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2cc17583e

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 3757–3759
www.rsc.org/chemcomm
COMMUNICATION
Fibrillisation of ring-closed amyloid peptidesw
a
Ian W. Hamley,*a Ge Cheng,z Valeria Castelletto,a Stephan Handschinb and
Raffaele Mezzengab
Received 5th December 2011, Accepted 21st February 2012
DOI: 10.1039/c2cc17583e
Ring-closing olefin metathesis reactions are used to create intra-
molecularly ring closed peptides or inter-molecularly ring-closed
peptide dimers based on a designed amyloid peptide sequence.
The uncrosslinked peptide self-assembles into high aspect ratio
nanotubes, however ring-closing leads to the formation of fibrillar
and twisted/helical ribbon structures.
studies on the fibrillisation of ring-closed peptides, and this is
discussed here for the first time.
The design of the peptide containing two allyl- functionalized
aspartic acid residues, D(All) is shown in Scheme 1. The peptide was
designed based on the KLVFFAE sequence Ab(16–22) from the
amyloid beta peptide.11–13 This contains the core KLVFF sequence
in amyloid fibrillisation and has been exploited as the basis for a
series of designed peptides under investigation in our group.14–18
The sequence was modified to EAFFD(All)D(All)VVLK,
peptide I. The use of the allyl-functionalized glutamic acid
residues, D(All), as cross-linkable residues, was motivated by
the commercial availability of the corresponding Fmoc-D(All)
amino acid. Peptide Ac-KLVFFAE-NH2 has been shown to self-
assemble into nanotubes in pH 2 acetonitrile/H2O solutions.12,19
Incorporating the two allyl ester aspartic acid residues as well
as an additional Val residue in peptide I (also with a reverse
sequence) still leads to nanotube formation, and we were
therefore able, uniquely, to investigate the influence of metathesis
ring-closing on peptide nanotube self-assembly.
The stabilization of peptide conformation is a major challenge
in the development of peptide therapeutics. Peptides offer
great potential in targeted delivery, however their stability
against proteolysis often needs improving. A number of
approaches have been developed to stabilize peptide and
protein conformation including stapling methods in which
intra-molecular cross-linking reactions are used to cross-link
peptides or sequences within proteins.1 For example, Grubbs
and coworkers used ring-closing olefin metathesis to cross-link
amino acid side chains to produce macrocyclic helical peptides
based on model hydrophobic sequences.2,3 The a-helix within
peptides based on a key domain of the p53 tumor suppressor
protein has also been stapled using olefin metathesis methods,4
as have proteins involved in the regulation of apoptosis5 and
peptides designed to bind at the coactivator binding site of the
estrogen receptor.6 Other methods have also been used to
covalently cross-link peptides including the application of the
‘‘click’’ copper-catalysed alkyne-azide cycloaddition reaction
to construct stabilized helices7,8 and b-sheets.9 These reactions
may also conveniently be used to prepare cyclic peptides.10
Here, we report on the olefin-metathesis cross-linking of
designed amyloid peptides which self-assemble into nanotubes
that comprise wrapped b-sheet walls. The cross-linking reaction
leads to intra-molecularly ring closed peptides, or inter-molecularly
ring-closed dimers. Despite the constraint imposed by the
architecture of the cross-linked fibrils, b-sheet fibril formation
is still observed. The width of these objects is controlled by the
nature of the ring-closing process. We are not aware of prior
The peptide E(OtBu)AFFD(All)D(All)VVLK(Boc) containing
cross-linkable allyl-functionalized aspartic acid residues
[D(All)] was prepared by solid phase peptide synthesis methods.
Deprotection afforded peptide I. Cross-linking reactions were
performed with the peptide on resin20,21 using an olefin metathesis
reaction using a second generation Hoveyda–Grubbs catalyst as
described in the Supporting Information. Two products, peptides
II and III were obtained from this peptide. Peptide II is an intra-
molecularly ring-closed peptide, while III is a ring-closed peptide
dimer. In addition, byproducts associated with the peptide I
lacking the A residue were also identified and labelled IV, V.
These are discussed further in the Supporting Information.
a Dept of Chemistry, University of Reading, Whiteknights Reading
RG6 6AD, UK. E-mail: I.W.Hamley@reading.ac.uk
b Food & Soft Materials Science, Institute of Food Nutrition &
Health, ETH Zurich, LFO23 Schmelzbergstrasse 9, 8092 Zurich,
Switzerland
¨
¨
w Electronic supplementary information (ESI) available: Experimental
Methods and characterization data. Additional XRD and TEM data,
and FTIR spectra. See DOI: 10.1039/c2cc17583e
z Present address: Dept of Chemistry, University of Liverpool, Crown
Street, Liverpool L69 3BX, UK.
Scheme 1 Structures of peptide I and alkene-bridged peptides II, III.
Chem. Commun., 2012, 48, 3757–3759 3757
c
This journal is The Royal Society of Chemistry 2012

View Online
to the horizontal flow direction. These maxima are the primary
form factor peaks in Fig. 1b.
The influence of cross-linking (intra-molecular in the case of II,
inter-molecular dimerization in the case of III) was examined by
cryo-TEM and SAXS. Fig. 2 presents cryo-TEM images. Peptide
II does not self-assemble at 0.5 wt% as shown by the absence of
structures in the cryo-TEM images (Fig. 2a and b). However, at
higher concentration (2 wt%) b-sheet formation is observed by
FTIR (ESI,w Fig. 11) and fibrils are noted in cryo-TEM images
(ESI,w Fig. 12). In contrast, cryo-TEM indicates that Peptide III
forms a mixture of straight fibrils and shorter irregular aggre-
gates (Fig. 2c and d) even at 0.5 wt%. This is supported by
FTIR (ESI,w Fig. 11) and negative stain TEM (ESI,w Fig. 13).
For the fibrils, the average diameter is (6.3 ꢀ 0.5) nm. The
cryo-TEM images therefore reveal that neither II nor III form
the nanotube structure of the uncross-linked parent peptide I.
To complement real-space imaging, and to provide an in situ
probe of nanostructure over a larger volume than cryo-TEM,
SAXS (and WAXS) was performed. Fig. 3 shows representative
data. In contrast to the data shown in Fig. 1, nanotube form
factor oscillations are not observed for solutions of either
peptide II or III (Fig. 3). These intensity profiles result from
fibril self-assembly—the form factor maximum at q = 0.6 nmꢁ1
(peptide III) is associated with the diameters of the fibrils shown
in Fig. 2.
Fig. 1 Peptide I forms nanotubes in aqueous solution. (a) Cryo-TEM
image obtained from a 2.4 wt% solution, (b) Integrated SAXS
intensity profile (black points) and WAXS profile (blue points) along
with calculated low q form factor (red line) measured for a 2 wt%
solution. A similar profile was observed for an 0.5 wt% solution,
(c) 2D-SAXS pattern from the same 2 wt% sample, showing alignment
of nanotubes along the (horizontal) flow direction.
In the following, we investigate the self-assembly of II and
III compared to the uncross-linked peptide I using cryogenic
transmission electron microscopy (cryo-TEM) and small angle
X-ray scattering (SAXS) to probe the nanostructure in aqueous
solution; X-ray diffraction is further employed to elucidate
details of the molecular packing. Unexpectedly, despite the steric
constraints induced by ring-closing, the formation of amyloid
fibrils is observed for II and III.
The WAXS data obtained from the solutions is particularly
interesting. In the WAXS profile for peptide III shown in
Fig. 3, peaks are observed at q = 5.6 nmꢁ1 (d = 1.1 nm), q =
13.1 nmꢁ1 (d = 0.48 nm) and q = 16.5 nmꢁ1 (shoulder,
d = 0.38 nm). Peptide II shows a broad shoulder at around
q = 13.5 nmꢁ1 (d = 0.47 nm). Consistent with cryo-TEM,
peptide II does not form well-developed b-sheets at low
concentration. However for peptide III the in situ X-ray
measurements, confirm the features of b-sheet ordering, i.e.
the spacing of the sheets (1.1 nm) and the strand spacing
(0.47–0.48 nm) within the b-sheets.
The self-assembly of peptide I into nanotubes in aqueous
solution was revealed by cryo-TEM and SAXS. Fig. 1a shows
a cryo-TEM image which reveals tube-like structures with a
radius (25 ꢀ 4) nm. The tube walls appear to be thin,
comprising B10% of the tube diameter. A nanotube structure
was confirmed by SAXS. Fig. 1b shows the SAXS intensity
profile, measured over an extended q range. The intensity
profile at low q contains oscillations from the form factor of
nanotubes. The red line shows a fit to an approximate form
factor of nanotubes with infinitely thin walls, and with a radius
of 22 nm, and a Gaussian polydispersity s = 10%. This form
factor captures the location of the form factor minima.
Further detailed modeling is prohibited by the presence of
an additional contribution to the scattering from structure
factor in the intermediate q range (giving rise to a broad
intensity maximum).
To further investigate nanostructure, X-ray diffraction was
performed on dried stalks. Representative images are shown
At high q, a series of Bragg peaks are observed that are due
to the configuration of the peptides within the nanotube walls
(to be discussed shortly). In addition, the WAXS profile
measured simultaneously (Fig. 1b and expanded in ESI,w
Fig. 10) contains peaks assigned to the b-sheet structure of
the peptides, specifically sharp peaks due to the b-strand
spacing (0.47 nm) and the b-sheet stacking distance (1.06 nm)
as well as several additional reflections. Spontaneous alignment
of peptide nanotubes upon transfer into the capillary flow-
through cell used for SAXS was observed, as revealed by the
two-dimensional SAXS profile in Fig. 1c. This shows intensity
maxima along the meridional (vertical) direction, i.e. perpendicular
Fig. 2 Cryo-TEM images. (a,b) Peptide II, 0.5 wt% in H2O, (c,d)
Peptide III, 0.25 wt% in H2O.
c
3758 Chem. Commun., 2012, 48, 3757–3759
This journal is The Royal Society of Chemistry 2012

View Online
can be estimated23 as 3.4 nm. The observed d = 1.8 nm peak
may be a second order reflection since it is highly unlikely this
peptide is folded.
In summary, ring-closing metathesis can be used to
create an intra-molecularly folded U-shaped peptide, whereas
inter-molecular ring-closing leads to extended a four-arm
‘‘branched’’ peptide dimer construct. The latter resembles in
some respects a cyclic peptide, although with a different
architecture (e.g. much larger branches than the side chains
on cyclic peptides) and with a distinct synthesis procedure.
Regions of these peptides must retain the b-sheet hydrogen
bonding pattern, consistent with the XRD results presented.
The conformational constraints introduced by ring-closing
hinder nanotube formation, observed for the parent uncross-
linked peptide. Instead, distinct fibrillar self-assembled nano-
structures are observed. The thickness of the fibril and ribbon
structures is related to the length of the peptides, whether
intra- or inter-molecularly ring-closed. Our results point towards
the use of ring-closing chemistries to create novel peptide
architectures, which are remarkably still able to form b-sheet
secondary structures.
Fig. 3 SAXS (and selected WAXS) data for peptide II and III in
0.5 wt% aqueous solutions. The inset shows an enlargement of the
WAXS profiles (in box) with linear intensity and q scales.
in ESI,w Fig. 14. ESI,w Fig. 15 shows typical intensity profiles
from equatorial sector averages for peptide I and peptide II.
This highlights certain features associated with the nanotube
structure of peptide I, in particular the presence of peaks at
d = 0.55 nm and d = 0.45 nm which are absent for peptide II,
the XRD pattern for which shows classical ‘‘cross-b’’ pattern
features, i.e. the 1.1 nm and 0.48 nm spacings (the same
features were observed for peptide III, data not shown). The
peaks observed for peptide I were previously observed in
oriented X-ray diffraction patterns (obtained by flow alignment)
for the heptapeptide A6K which forms nanotubes in concentrated
aqueous solution.22 A model for the origin of these peaks was
presented based on the helical wrapping of peptide dimers
(with a two-residue offset between adjacent dimers). An additional
feature noted in the fibre XRD is a d = 1.8 nm peak for both
peptides. For peptide I, it may be noted that a strong meridional
reflection presumably associated with the b-strand spacing, is
shifted such that the spacing is larger than usual, d = 0.49 nm.
The d = 0.45 nm peak is only present as a shoulder. The
d = 0.39 nm peak for peptide I is sharper and more intense than
the corresponding d = 0.38 nm peak observed for peptide II.
This peak is associated with the Ca–Ca backbone spacing.22
Molecular modeling (Cerius Molecular Dynamics using a
DREIDING force field) suggests a favored hairpin confor-
mation for peptide II, with a length approximately 1.8 nm
(Fig. 4a). This is in excellent agreement with the d spacing
observed in the fibre XRD pattern. The thickness of fibrils
observed by cryo-TEM suggests a width of three strands.
For peptide III, molecular dynamics simulations lead to a
conformation (Fig. 4b) with a planar arrangement in the cross-
linked region, and with a longest dimension approximately
3.5–4 nm. Detailed modeling of the aggregation of the peptides
into b-sheet fibrils will be the subject of future work. The length
of the decapeptide I in an antiparallel b-sheet configuration
Notes and references
1 Y. Y. A. Lin, J. M. Chalker and B. G. Davis, ChemBioChem, 2009,
10, 959.
2 H. E. Blackwell and R. H. Grubbs, Angew. Chem., Int. Ed., 1998,
37, 3281.
3 H. E. Blackwell, J. D. Sadowsky and R. J. Howard, et al., J. Org.
Chem., 2001, 66, 5291.
4 F. Bernal, A. F. Tyler and S. J. Korsmeyer, et al., J. Am. Chem.
Soc., 2007, 129, 2456.
5 L. D. Walensky, A. L. Kung and I. Escher, et al., Science, 2004,
305, 1466.
6 C. Phillips, L. R. Roberts and M. Schade, et al., J. Am. Chem. Soc.,
2011, 133, 9696.
7 S. Cantel, A. L. C. Isaad and M. Scrima, et al., J. Org. Chem.,
2008, 73, 5663.
8 O. Jacobsen, H. Maekawa and N. H. Ge, et al., J. Org. Chem.,
2011, 76, 1228.
9 T.-B. Yu, J. Z. Bai and Z. Guan, Angew. Chem., Int. Ed., 2009,
48, 1097.
10 U. Kazmaier, C. Hebach and A. Watzke, et al., Org. Biomol.
Chem., 2005, 3, 136.
11 I. W. Hamley, Angew. Chem., Int. Ed., 2007, 46, 8128.
12 K. Lu, J. Jacob and P. Thiyagarajan, et al., J. Am. Chem. Soc.,
2003, 125, 6391.
13 Y. Liang, S. V. Pingali and A. S. Jogalekar, et al., Biochemistry,
2008, 47, 10018.
14 M. J. Krysmann, V. Castelletto and A. Kelarakis, et al., Biochemistry,
2008, 47, 4597.
15 M. J. Krysmann, V. Castelletto and J. M. E. McKendrick, et al.,
Langmuir, 2008, 24, 8158.
16 V. Castelletto, I. W. Hamley and R. A. Hule, et al., Angew. Chem.,
Int. Ed. Engl., 2009, 48, 2317.
17 V. Castelletto, I. W. Hamley and C. Cenker, et al., J. Phys.
Chem. B, 2011, 115, 2107.
18 G. Cheng, C. Krasel and H. G. Zhou, et al., Org. Lett., 2011,
13, 2572.
19 A. K. Mehta, K. Lu and W. S. Childers, et al., J. Am. Chem. Soc.,
2008, 130, 9829.
20 S. J. Miller, H. E. Blackwell and R. H. Grubbs, J. Am. Chem. Soc.,
1996, 118, 9606.
21 G. Dimartino, D. Y. Wang and R. N. Chapman, et al., Org. Lett.,
2005, 7, 2389.
22 V. Castelletto, D. R. Nutt and I. W. Hamley, et al., Chem.
Commun., 2010, 46, 6270.
23 T. E. Creighton, ‘Proteins. Structures and Molecular Properties’,
W.H. Freeman, 1993.
Fig. 4 Simulated conformations for (a) peptide II, (b) peptide III.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 3757–3759 3759