Tripeptide self-assembled hydrogels: Unexpected twists of chirality

Article abstract of DOI:10.1039/c2cc16609g

Change of chirality of the first N-terminal amino acid of tripeptides VFF and FFV from l to d results in self-assembled hydrogels at physiological pH from non-assembling l-analogues. Interestingly, changing the chirality of F yields very different nanostructures; nanotapes are observed for ^DVFF, twisted fibers for ^DFFV.

Full text of DOI:10.1039/c2cc16609g

View Article Online / Journal Homepage / Table of Contents for this issue
Chemical Communications
www.rsc.org/chemcomm
Volume 48 | Number 16 | 21 February 2012 | Pages 2157–2256
ISSN 1359-7345
COMMUNICATION
Silvia Marchesan et al.
Tripeptide self-assembled hydrogels: unexpected twists of chirality

Dynamic ArticVleiewLiAnrktiscle Online
ChemComm
Cite this: Chem. Commun., 2012, 48, 2195–2197
www.rsc.org/chemcomm
COMMUNICATION
Tripeptide self-assembled hydrogels: unexpected twists of chiralityw
Silvia Marchesan,*a Christopher D. Easton,a Firdawosia Kushkaki,b Lynne Waddingtona and
Patrick G. Hartleya
Received 25th October 2011, Accepted 24th November 2011
DOI: 10.1039/c2cc16609g
Change of chirality of the first N-terminal amino acid of
tripeptides VFF and FFV from L to D results in self-assembled
hydrogels at physiological pH from non-assembling L-analogues.
Interestingly, changing the chirality of F yields very different
nanostructures; nanotapes are observed for DVFF, twisted fibers
for DFFV.
The self-assembly behaviour of amyloid peptides is the topic of
intense research due to both pathological relevance and
Fig. 1 Dynamic time sweep rheological data of freshly made hydrogels
potential application in the production of nanostructured
materials.1 In particular, the core sequence KLVFF has been
identified as the shortest key fragment responsible for the
formation of b-sheet fibrils, which can yield hydrogels.2 At
present, there are no reported examples of KLVFF fragments
(left). The hydrogels formed are self-supporting (right).
chirality in the case of DFFV results in highly twisted nano-
structures, and CD, FT-IR and Thioflavin T fluorescence
confirm they consist of b-sheets.
capable of self-assembly and hydrogelation in the absence of
organic solvents: uncapped di- and tripeptides are either too
hydrophilic and readily dissolve in water, or too hydrophobic, and
thus precipitate. In the case of FF alone, an elegant study revealed
discrete nanotubes of persistent length, when a concentrated
solution of the dipeptide in hexafluoroisopropanol is diluted
in water.3 Later, a few examples were reported of capped
F derivatives, including di- and tripeptides, capable of self-assembly
when first dissolved at high concentration in an organic solvent.4
This is not surprising as it is known that the self-assembly process is
strongly dependent on solvent.5 Different is the case for Fmoc-FF
and naphthalene-FF, that do not require the presence of organic
solvents to hydrogelate, but where self-assembly is driven by the
synthetic aromatic moiety.6
In this study, we used a pH switch to promote self-assembly:
peptides were first dissolved in a 0.1 M sodium phosphate
solution at pH 12, while subsequent dilution to final physiological
pH at B7.4 triggered the formation of macroscopic hydrogels
only for DVFF and DFFV, but not for their L-analogues, as
confirmed by visual observation and rheology (Fig. 1). While
both gels were initially very soft (G0 of 10 kPa was reached only
after 1 hour for both peptides), it is worth noting that gelation of
DFFV could be markedly accelerated by sonication (see ESIw).
CryoTEM (Fig. 2) of the tripeptides containing a D-amino acid
showed nanoscale fibrous morphologies, whilst no structures
could be identified for their L-analogues. Interestingly, the resulting
nanostructures of the two non-natural peptides were markedly
different from each other. DVFF yielded long nanotapes, often
running in parallel (Fig. 2A, white arrowheads) and displaying
lateral stacking into wider structures (Fig. 2A, black arrowheads).
DFFV self-assembled into thicker filaments, usually running in
twisted pairs with a recurrent helix period of B150–200 nm
(Fig. 2B and C). These nanostructures were confirmed by
AFM on dried gel samples (Fig. 3A–D). In particular, a helix
period of B150–200 nm is compatible with the AFM linescan
of DFFV (line drawn in Fig. 3B and analysis in Fig. 3C) and is
shown in the image in Fig. 3D. Occasionally, multiple DFFV
filaments could also twist into coiled fibrils (see ESIw). In further
experiments, the precursor solutions were dried onto a silicon wafer
before gelation without sonication, in an attempt to gain insights
into the initial phases of self-assembly. Interestingly, DVFF yielded
similar structures to those observed after sonication (compare
Fig. 3A and E), while DFFV formed short nanorods which
suggest partial, short order alignment (Fig. 3F). It is possible
Here, we report that while the short tripeptides VFF and FFV
indeed do not self-assemble into fibrils in phosphate buffer, it is
sufficient to change the configuration from L to D of the N-terminal
amino acid in both sequences to obtain self-supporting hydrogels.
To the best of our knowledge, these are the first examples of
uncapped FF-containing tripeptides capable of self-assembly into
hydrogels at physiological pH without the aid of organic solvents.
Interestingly, Cryo-TEM and AFM data show that the change of
a CSIRO Materials Science and Engineering, Private Bag 10,
Bayview Avenue, Clayton VIC 3168, Australia.
E-mail: silvia.marchesan@csiro.au
b La Trobe University, Department of Chemistry, La Trobe Institute
of Molecular Sciences, Kingsbury Drive, Bundoora,
Melbourne VIC 3086, Australia
w Electronic supplementary information (ESI) available: Full experi-
mental details, including peptide synthesis and characterization;
further rheology, AFM, and TEM data. See DOI: 10.1039/c2cc16609g
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 2195–2197 2195

View Article Online
Fig. 4 Confocal images of Thioflavin T fluorescence highlight fibers
of similar diameter for DVFF (A) and DFFV (B); the latter yields a
denser network.
the natural peptides (see ESIw), in agreement with Cryo-TEM.
It is worth noting how the different behaviour (self-assembling
or not) and morphology (i.e. tapes versus twisted fibrils) results
from very subtle stereochemical changes to tripeptide
composition.
Thioflavin T fluorescence is a well-known method to assess
b-sheet structures, although it is only recently that a binding
mode has been proposed suggesting that the dye binds to an
aromatic–hydrophobic groove spanning across at least four
consecutive b-strands.7 Confocal imaging (Fig. 4) of our
samples treated with Thioflavin T revealed fibers of several
mms in length for both DVFF and DFFV, with the latter
yielding a denser network, as expected from our data. In
contrast, the lack of fluorescence for the two natural analogues
suggests they are devoid of extended b-sheets (not shown).
CD was used to better assess the nature of the secondary
structure of the peptide self-assembled materials (Fig. 5).
Surprisingly, the CD spectra of all four peptides showed an
evident maximum centered at 225–230 nm, indicative of p–p
stacking of the aromatic phenylalanine residues. It is plausible
that this is due to the formation of aggregates for the two
L-peptides, as their limited solubility led to slow precipitation.
It is worth noting that in the case of the self-assembling
peptide DFFV, the intensity of the maximum was significantly
reduced, due to the presence of the D-aromatic amino acid.
More importantly, as expected only the two self-assembling
structures containing non-natural amino acids displayed the
b-sheet minimum in the region 210–220 nm (Fig. 5, bottom).
The DFFV self-assembled peptide was the only one showing a
broad positive signal in the near-UV region (250–270 nm),
which can be attributed to phenylalanine residues with tertiary
structure. The b-sheet nature of the self-assembling peptides
Fig. 2 (A) CryoTEM of DVFF hydrogel. Black arrowheads: wide
nanotapes. White arrowheads: parallel double filaments. (B) CryoTEM
of DFFV hydrogel. Black arrowheads: helix period, more clearly visible
in the higher magnification image C. Scale bar = 200 nm.
Fig. 3 AFM images of dried gels of DVFF (A) and DFFV (B) obtained
after sonication; an AFM linescan (C) and a detailed view (D) of a helix
period of B150–200 nm for DFFV. Precursor solutions which were not
sonicated and were dried before gelation display similar structures for
DVFF (E) but only short nanorods for DFFV (F).
that these precursors eventually elongate into the twisted
fibrils observed for DFFV after sonication (Fig. 3B and ESIw).
In contrast, AFM did not reveal ordered nanostructures for
Fig. 5 CD spectra of the four tripeptides: each sample displays p–p
stacking (maximum at 225–230 nm); only the two self-assembling hydrogels
(bottom) display the b-sheet minimum in the region 210–220 nm.
c
2196 Chem. Commun., 2012, 48, 2195–2197
This journal is The Royal Society of Chemistry 2012

View Article Online
(Cryo-TEM, AFM and confocal) and spectroscopic (CD, FT-IR)
techniques. In addition, these gels are promising candidates
for the development of biomaterials as they form at physiolo-
gical pH without the aid of organic solvents, or of synthetic
capping end-groups, and the display of a D-amino acid at their
N-terminus may be a convenient feature to extend their
resistance to protease degradation.3 Investigation of the struc-
tural properties of a series of analogues is currently underway in
our laboratories to further elucidate how the chiral features
described in this study influence intermolecular interactions
and translate into the observed dramatic effects on the supra-
molecular structures. It can be envisaged that chirality will be a
key tool to control self-assembly behavior and also to induce
designed morphologies in nanostructured materials.
Fig. 6 FT-IR of the amide I–II regions for the four tripeptides.
was also confirmed by FT-IR spectral analysis of the amide
I–II regions (Fig. 6). The natural peptide VFF revealed a
mixture of secondary structures, including a predominant
maximum at 1610–1620 cmꢀ1 that can be attributed to antiparallel
aggregates, and which is absent or barely visible in all the other
samples. In comparison, the spectrum of FFV is very different,
with a broad maximum centered at 1640 cmꢀ1 ascribed mainly
to random coils, and a dominant TFA peak in the region
1670–1680 cmꢀ1 that makes it difficult to assess the presence of
b-structures. Nevertheless, both VFF and FFV are likely devoid
of extended b-sheets with supramolecular order, as suggested by
CD analysis and lack of binding and fluorescence of Thioflavin T.
TFA presence was apparent also in the spectra of the two
non-natural self-assembling peptides, however, in this case
the antiparallel b-sheet signature was evident at least for
DFFV (1637 and 1680–1690 cmꢀ1).
Moreover, DVFF and DFFV both displayed a broad signal
in the N–H amide II region, shifted to lower frequencies due to
extended hydrogen bonding, indicative once again of b-sheets.
More specifically, the signal at 1570 cmꢀ1 suggests COOꢀ/
NH3+ head-to-tail interactions in the self-assembling tripeptides,
as previously observed for the gel-forming Ile-Phe but not for
the non-self-assembling Val-Phe.4d It is interesting to note that
a maximum centered at 1523 cmꢀ1 was present in the spectrum
of VFF, but not FFV. It appears that the overall lack of
supramolecular order for the two natural peptides can be
ascribed to the co-existence of a number of secondary structures
for VFF, and their absence for FFV, which appears to be
predominantly random coil.
The authors acknowledge the facilities of Monash Micro
Imaging, Monash University, Australia, and in particular
Stephen Firth, Dr Judy Callaghan and Dr Alex Fulcher for
their scientific and technical assistance.
Notes and references
1 (a) T. R. Jahn, O. S. Makin, K. L. Morris, K. E. Marshall, P. Tian,
P. Sikorski and C. Serpell, J. Mol. Biol., 2010, 395(4), 717;
(b) I. Cherny and E. Gazit, Angew. Chem., Int. Ed., 2008,
47(22), 4062; (c) K. Channon and C. E. MacPhee, Soft Matter,
2008, 4(4), 647.
2 (a) L. O. Tjernberg, J. Naoslund, F. Lindqvist, J. Johansson,
A. R. Karlstroem, J. Thyberg, L. Terenius and C. Nordstedt,
J. Biol. Chem., 1996, 271(15), 8545; (b) M. J. Krysmann,
V. Castelletto, A. Kelarakis, I. W. Hamley, R. A. Hule and
D. Pochan, Biochemistry, 2008, 47(16), 4597.
3 M. Reches and E. Gazit, Science, 2003, 300, 625.
4 (a) D. M. Ryan, T. M. Doran and B. L. Nilsson, Langmuir, 2011,
27(17), 11145; (b) D. Haldar, S. Maity and P. Kumar, Org. Biomol.
Chem., 2011, 9(10), 3787; (c) P. Tamamis, L. Adler-Abramovich,
M. Reches, K. Marshall, P. Sikorski, L. Serpell, E. Gazit and
G. Archontis, Biophys. J., 2009, 96, 5020; (d) N. S. de Groot,
T. Parella, F. X. Aviles, J. Vendrell and S. Ventura, Biophys. J.,
2007, 92(5), 1732; (e) X. H. Yan, Q. He, K. W. Wang, L. Duan,
Y. Cui and J. B. Li, Angew. Chem., Int. Ed., 2007, 46(14), 2431.
5 (a) P. Jonkheijm, P. van der Schoot, A. P. H. Schenning and
E. W. Meijer, Science, 2006, 313, 80; (b) V. Castelletto,
I. W. Hamley, P. J. F. Harris, U. Olsson and N. Spencer, J. Phys.
Chem. B, 2009, 113(29), 9978.
6 (a) L. Chen, G. Pont, K. Morris, G. Lotze, A. Squires, L. C.
Serpell and D. J. Adams, Chem. Commun., 2011, 47, 12071;
(b) D. J. Adams, Macromol. Biosci., 2011, 11, 160; (c) L. Chen,
S. Revel, K. Morris, L. C. Serpell and D. J. Adams, Langmuir, 2010,
26, 13466; (d) V. Jayawarna, A. Murtza, T. A. Jowitt, A. F. Miller,
A. Saiani, J. E. Gough and R. V. Ulijn, Adv. Mater., 2006,
18(5), 611.
In conclusion, this study reports two examples where
chirality acts as a key tool for the self-assembly of short
peptides. This effect is not limited to isolated nanostructures,
but extends to the formation of macroscopic self-supporting
hydrogels, as confirmed by rheology and a variety of microscopic
7 J. E. Shea, C. Wu, M. Biancalana and S. Koide, J. Mol. Biol., 2009,
394(4), 627.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 2195–2197 2197