A reductive trigger for peptide self-assembly and hydrogelation

Article abstract of DOI:10.1021/ja1025535

Stimulus-responsive peptide self-assembly provides a powerful method for controlling self-assembly as a function of environment. The development of a reductive trigger for peptide self-assembly and subsequent hydrogelation is described herein. A self-assembling peptide sequence, Ac-C(FKFE) ₂CG-NH₂, was cyclized via disulfide bonding of the flanking cysteine residues. The macrocyclic form of this peptide enforces a conformational restraint that prevents adoption of the β-sheet conformation that is required for self-assembly. Upon reduction of this disulfide bond, the peptide relaxes into the preferred β-sheet conformation, and immediate self-assembly into fibrillar superstructures occurs. At sufficient peptide concentration, self-assembly is accompanied by the formation of rigid, viscoelastic hydrogels.

Full text of DOI:10.1021/ja1025535

Published on Web 04/21/2010
A Reductive Trigger for Peptide Self-Assembly and Hydrogelation
Charles J. Bowerman and Bradley L. Nilsson*
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627-0216
Received March 26, 2010; E-mail: nilsson@chem.rochester.edu
Scheme 1. Cyclic to Linear Peptide Conformational Switch Using a
Reductive Trigger
Abstract: Stimulus-responsive peptide self-assembly provides a
powerful method for controlling self-assembly as a function of
environment. The development of a reductive trigger for peptide
self-assembly and subsequent hydrogelation is described herein.
A self-assembling peptide sequence, Ac-C(FKFE)2CG-NH2, was
cyclized via disulfide bonding of the flanking cysteine residues.
The macrocyclic form of this peptide enforces a conformational
restraint that prevents adoption of the ꢀ-sheet conformation that
is required for self-assembly. Upon reduction of this disulfide
bond, the peptide relaxes into the preferred ꢀ-sheet conformation,
and immediate self-assembly into fibrillar superstructures occurs.
At sufficient peptide concentration, self-assembly is accompanied
by the formation of rigid, viscoelastic hydrogels.
like fibrillar bilayers, was chosen as a model sequence to explore
the concept of a reductive trigger for peptide self-assembly.13,14
This sequence was flanked with Cys residues, and a Gly was
included at the C-terminus to avoid problems with Cys racemization
during resin loading; this gave the target sequence, Ac-C(FKFE)2CG-
NH2 (linear-1, Scheme 1). Linear-1 was synthesized by standard
Fmoc solid-phase peptide synthesis and then oxidatively cyclized
by diluting linear-1 (e100 µM, acetonitrile/water) and incubating
it with 4,4′-pyridyldisulfide (4-PDS) for 24 h, giving cyclic-1.15
The presence of high ratios of acetonitrile prevented self-assembly
of linear-1, thus facilitating these operations. HPLC analysis
indicated nearly quantitative conversion of linear-1 to cyclic-1 under
these conditions.
Noncovalent peptide self-assembly has been exploited for the
preparation of dynamic and functional hydrogel nanomaterials that
have been applied to regenerative medicine, wound healing, drug
delivery, and biosensing.1-5 Upon undergoing self-assembly at
sufficient concentration, peptide-based fibrils form entangled net-
works that reduce the flow of solvent, resulting in the hydrogel
state.
Stimulus-responsive peptide self-assembly provides a powerful
method for controlling hydrogelation as a function of environment.6
Peptides have been designed to undergo self-assembly in response
to environmental and chemical stimuli,4 including pH,7 ionic
strength,8 temperature,9 light,10 and enzymatic manipulation.11 Of
these stimuli-responsive systems, only ones based on pH or enzyme-
triggered self-assembly are highly responsive to biological mi-
croenvironments. For example, self-assembly can be triggered
within a cell by choosing an enzyme activator that is not found in
the extracellular environment. The development of additional
microenvironment-sensitive triggers for self-assembly will enable
more diverse biological application of peptide-derived self-as-
sembled materials.
Figure 1. (A) CD spectra of cyclic-1 (1 mM in water) and cyclic-1 +
TCEP (giving linear-1; 1 mM peptide, 10 mM TCEP). (B) Negatively
stained TEM image of fibrils derived from linear-1 (cyclic-1 + TCEP).
Imposing a conformational constraint that prevents adoption of
the ꢀ-sheet secondary structure required for the self-assembly of
short peptides is a possible strategy for controlling peptide self-
assembly. We postulated that flanking a short self-assembling
peptide sequence with cysteine (Cys) residues would enable the
macrocyclization of these peptides, preventing ꢀ-sheet formation
and self-assembly in the cyclic form (Scheme 1). This constraint
could be removed by simple reduction of the disulfide bond,
resulting in relaxation to the stable ꢀ-strand and thus triggering
self-assembly. A reductive trigger could facilitate more nuanced
stimulus-responsive self-assembly in a biological milieu. For
example, a peptide could remain monomeric in the extracellular
oxidizing environment but assemble in the reducing environment
of the cell or in the vicinity of tumors, which often have local
extracellular reducing environments.12
The self-assembly behavior of the cyclic-1 peptide was then
assessed. Self-assembly studies were performed by dissolving the
peptides in unbuffered water [pH 2.4-3.0 because of the presence
of residual trifluoroacetic acid (TFA) from HPLC purification].
Circular dichroism (CD) spectra of cyclic-1 at concentrations
ranging from 100 µM to 10 mM showed a characteristic minimum
at 208 nm indicative of a random-coil structure (Figure 1A and
Figure S2 in Supporting Information). Negatively stained transmis-
sion electron microscopy (TEM) images of solutions of cyclic-1
showed no evidence of fibrillar self-assembly. Solutions of cyclic-1
were stable for months, with no apparent self-assembly even after
extended incubation at all concentrations studied (Figure S9).
The addition of reductant to cyclic-1 resulted in immediate
disulfide bond reduction and self-assembly of the resulting linear-1
The amphipathic Ac-(FKFE)2-NH2 sequence, which has a high
propensity to undergo self-assembly into water-soluble amyloid-
9
9526 J. AM. CHEM. SOC. 2010, 132, 9526–9527
10.1021/ja1025535  2010 American Chemical Society

C O M M U N I C A T I O N S
peptide. Tris(2-carboxyethyl)phosphine (TCEP) was initially chosen
as the reductant, and 1.2-10 molar equiv relative to disulfide was
added. Dithiothreitol (DTT) was also an effective reductant,
triggering formation of linear-1 and self-assembly at the same
concentrations as TCEP. The CD spectrum of cyclic-1 + TCEP
displays two major minima at 217 and 200 nm (Figure 1A). The
minimum at 217 nm is classically indicative of the ꢀ-sheet structure
required for peptide self-assembly of the (FKFE)2 sequence. The
minimum at 200 nm has been observed previously with the (FKFE)2
peptide and attributed to π-π* effects resulting from aromatic π-π
interactions in the hydrophobic interior of the self-assembled bilayer
structure.14,16 Negatively stained TEM images of linear-1 solutions
displayed abundant peptide fibrils that were 3.5 ( 0.4 nm wide
(Figure 1B). Self-assembly was observed at all concentrations
studied (100 µM to 10 mM; see Figures S3-S8 for CD spectra
and TEM images). This data indicates rapid peptide self-assembly
upon disulfide bond reduction and relaxation of conformational
constraint.
after maturation for 24 h (Figure S12) showed only subtle
differences in fibril density, suggesting that intrasheet or intersheet
bilayer disulfide bonds dominate. Increasing the temperature also
resulted in rigidification of the hydrogels, as has been observed in
other systems (at 75 °C, G′ ) 3386.9 ( 87.4 Pa and G′′ ) 326.2
( 15.6 Pa; see the Supporting Information for temperature-sweep
data).9
This work has demonstrated that introduction of a conformational
constraint by disulfide-bond-mediated peptide cyclization is a viable
strategy for controlling peptide self-assembly. This strategy should
be generally applicable to many short self-assembling peptide
sequences. Significantly, reducing and oxidizing conditions are
important in biologically relevant microenvironments, and the
development of a reductive trigger for stimulus-responsive self-
assembly has considerable potential for application in biotechnol-
ogy. Exploration of potential applications is currently underway in
our lab. Finally, use of conformational constraints to control peptide
self-assembly need not be limited to reductive triggers. Many
reversible chemical bonds could potentially be utilized to the same
effect by relying on other environmental stimuli to promote self-
assembly, and the realization of this potential will open novel and
imaginative opportunities for the exploitation of peptide self-
assembly in biomaterials and biomedical research.
Hydrogelation was observed in a concentration-dependent man-
ner upon addition of reductant to solutions of cyclic-1. At low
concentrations (e1 mM), linear-1 solutions did not form hydrogels.
At a peptide concentration of 10 mM (∼0.9 wt %), addition of
TCEP resulted in the immediate formation of self-supporting
hydrogels, as indicated by the vial inversion test (Figure 2A). The
viscoelastic properties of these hydrogels were characterized
rheologically: the storage modulus (G′) and loss modulus (G′′)
indicate the elasticity and viscosity of the hydrogel, respectively
(Figure 2B). A rigid hydrogel is defined in practice as having a G′
value that exceeds the G′′ value by an order of magnitude.17
Immediately after addition of reductant, linear-1-derived hydrogels
exhibited G′ ) 79.9 ( 1.3 Pa and G′′ ) 7.9 ( 0.8 Pa. The resulting
hydrogels were stable at oscillatory frequencies of up to 30 rad
Acknowledgment. We acknowledge Dr. Scott Kennedy and
Karen Bentley (URMC Electron Microscope Research Core) for
assistance with CD and TEM experiments. We thank Chris
Willoughby and TA Instruments for the use of an AR-G2 rheometer.
This work was supported by a Department of Education GAANN
Fellowship to C.J.B., a DuPont Young Professor Award to B.L.N.,
the Alzheimer’s Association (NIRG-08-90797), and ACS PRF
(48922-DNI1).
s-1
.
Supporting Information Available: Experimental details and
additional CD spectroscopic data and TEM images. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Cavalli, S.; Albericio, F.; Kros, A. Chem. Soc. ReV. 2010, 39, 241.
(2) Kyle, S.; Aggeli, A.; Ingham, E.; McPherson, M. Trends Biotechnol. 2009,
27, 423.
(3) Jung, J. P.; Gasiorowski, J. Z.; Collier, J. H. Biopolymers 2010, 94, 49.
(4) Cui, H.; Webber, M. J.; Stupp, S. I. Biopolymers 2010, 94, 1.
(5) Branco, M.; Schneider, J. P. Acta Biomater. 2009, 5, 817.
(6) Mart, R. J.; Osborne, R. D.; Stevens, M. M.; Ulijn, R. V. Soft Matter 2006,
2, 822.
(7) (a) Rajagopal, K.; Lamm, M. S.; Haines-Butterick, L. A.; Pochan, D. J.;
Schneider, J. P. Biomacromolecules 2009, 10, 2619. (b) Aggeli, A.; Bell,
M.; Carrick, L. M.; Fishwick, C. W. G.; Harding, R.; Mawer, P. J.; Radford,
S. E.; Strong, A. E.; Boden, N. J. Am. Chem. Soc. 2003, 125, 9619.
(8) (a) Ozbas, B.; Kretsinger, J.; Rajagopal, K.; Schneider, J. P.; Pochan, D. J.
Macromolecules 2004, 37, 7331. (b) Carrick, L. M.; Aggeli, A.; Boden,
N.; Fisher, J.; Ingham, E.; Waigh, T. A. Tetrahedron 2007, 63, 7457.
(9) Pochan, D. J.; Schneider, J. P.; Kretsinger, J.; Ozbas, B.; Rajagopal, K.;
Haines, L. J. Am. Chem. Soc. 2003, 125, 11802.
Figure 2. (A) Images of vial inversion tests for (left) cyclic-1 (vial upright
and inverted, white arrows indicate the liquid levels) and (right) linear-1
(i.e., cyclic-1 + TCEP; vial upright and inverted). (B) Rheological storage
moduli (G′) and loss moduli (G′′) of hydrogels derived from linear-1 (i.e.,
cyclic-1 + TCEP) at 25 and 75 °C.
(10) Haines, L. A.; Rajagopal, K.; Ozbas, B.; Salick, D. A.; Pochan, D. J.;
Schneider, J. P. J. Am. Chem. Soc. 2005, 127, 17025.
(11) Gao, Y.; Yang, Z.; Kuang, Y.; Ma, M.; Li, J.; Zhao, F.; Xu, B. Biopolymers
2010, 94, 19.
Rigidification of the hydrogel was observed when the gel was
allowed to mature for 24 h. The G′ and G′′ values increased to
160.2 ( 14.5 and 15.8 ( 1.3 Pa, respectively, and these gels were
stable to 50 rad s-1. This is most likely due to the formation of
covalent disulfide cross-links between fibrils (or between ꢀ-strands
within a ꢀ-sheet or between ꢀ-sheets in the bilayer) as reductant
was consumed and an oxidizing environment subsequently
established.18,19 Disulfide bond formation was indicated by deter-
mining the free thiol concentration using Ellman’s reagent.15 The
thiol concentration was 8.2 mM immediately after addition of TCEP
reductant. After 24 h, this concentration was reduced to 2.3 mM,
consistent with thiol oxidation leading to hydrogel stabilization and
rigidification. TEM images of the freshly prepared gel and the gel
(12) Aluri, S.; Janib, S. M.; Mackay, J. A. AdV. Drug DeliVery ReV. 2009, 61,
940.
(13) Caplan, M. R.; Moore, P. N.; Zhang, S.; Kamm, R. D.; Lauffenburger,
D. A. Biomacromolecules 2000, 1, 627.
(14) Bowerman, C. J.; Ryan, D. M.; Nissan, D.; Nilsson, B. L. Mol. BioSyst.
2009, 5, 1058.
(15) Cline, D. J.; Thorpe, C.; Schneider, J. P. Anal. Biochem. 2004, 335, 168.
(16) Wang, K.; Keasling, J. D.; Muller, S. J. Int. J. Biol. Macromol. 2005, 261,
3.
(17) Menger, F. M.; Caran, K. L. J. Am. Chem. Soc. 2000, 122, 11679.
(18) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 5133.
(19) Aulisa, L.; Dong, H.; Hartgerink, J. D. Biomacromolecules 2009, 10, 2694.
JA1025535
9
J. AM. CHEM. SOC. VOL. 132, NO. 28, 2010 9527