Tailoring Supramolecular Peptide–Poly(ethylene glycol) Hydrogels by

May 24, 2016 - Melis Goktas , Chuanfu Luo , Ruby May A. Sullan , Ana E. Bergues-Pupo , Reinhard Lipowsky , Ana Vila Verde , Kerstin G. Blank. Chemical...
0 downloads 0 Views 2MB Size
Subscriber access provided by UCL Library Services

Article

Tailoring Supramolecular Peptide-Poly(ethylene glycol) Hydrogels by Coiled Coil Self-Assembly and Self-Sorting Staffan Danmark, Christopher Aronsson, and Daniel Aili Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.6b00528 • Publication Date (Web): 24 May 2016 Downloaded from http://pubs.acs.org on May 25, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biomacromolecules is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Tailoring

Supramolecular

Peptide-Poly(ethylene

glycol)

Hydrogels by Coiled Coil Self-Assembly and Self-Sorting Staffan Dånmark‡, Christopher Aronsson‡, and Daniel Aili*

Division of Molecular Physics, Department of Physics, Chemistry and Biology, Linköping University, 581 83, Linköping, Sweden ‡

These authors contributed equally to this work

Abstract Physical hydrogels are extensively used in a wide range of biomedical applications. However, different applications require hydrogels with different mechanical and structural properties. Tailoring these properties demands exquisite control over the supramolecular interactions involved. Here we show that it is possible to control the mechanical properties of hydrogels using de novo designed coiled coil peptides with different affinities for dimerization. Four different non-orthogonal peptides, designed to fold into four different coiled coil heterodimers with dissociation constants spanning from µM to pM, were conjugated to star-shaped 4-armpoly(ethylene glycol) (PEG). The different PEG-coiled coil conjugates self-assemble as a result of peptide heterodimerization. Different combinations of PEG-peptide conjugates assemble into PEG-peptide networks and hydrogels with distinctly different thermal stabilities, supramolecular and rheological properties, reflecting the peptide dimer affinities. We also demonstrate that it is possible to rationally modulate the self-assembly process by means of thermodynamic self-sorting by sequential additions of non-pegylated peptides. The specific interactions involved in peptide dimerization thus provides means for programmable and reversible self-assembly of hydrogels with precise control over rheological properties,

1 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

which can significantly facilitate optimization of their overall performance and adaption to different processing requirements and applications.

Introduction Hydrogels are highly hydrated polymeric networks that are widely used in various biomedical applications, including 3D cell culture, cell therapy, and bioprinting.1,2 Hydrogels are attractive as cell carriers in injection-based therapies and in bioinks for 3D bioprinting as they can protect cells from damaging shear forces during the injection and the additive manufacturing process. Hydrogels can also offer an extracellular matrix mimicking environment that supports cell growth and differentiation, which is vital in both tissue engineering applications and 3D cell culture.3-5 Notably, different applications and cell types require hydrogels with significantly different properties. The influence of the mechanical, chemical and structural properties of the hydrogels on cell behavior, stem cell differentiation, tissue integration and ability to stimulate regeneration cannot be underestimated and these properties must thus be tightly controlled.6 Hydrogels must also withstand various forms of processing without loss of functionality meaning that they must be structurally robust at the molecular level and display tunable shear thinning and preferably self-healing properties. Although rheological properties of chemically cross-linked hydrogels can be varied using different cross-linking strategies, their synthesis can damage cells, biomolecules and surrounding tissue. In contrast, physical hydrogels can be obtained as a result of supramolecular interactions that typically are not harmful to cells and tissues, making them ideal for biomedical applications. Unfortunately, it is more difficult to rationally tune the rheological properties of physical hydrogels as it requires precise means to control the supramolecular interactions involved in the self-assembly process.

2 ACS Paragon Plus Environment

Page 2 of 24

Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

The enormous chemical and structural diversity of peptides make them ideal as supramolecular building blocks for bottom-up fabrication of materials. The increasing knowledge about sequence-to-structure relationships has facilitated design of peptides that adopt well-defined structural motifs, such as coiled coils, helix-loop-helices and beta-hairpins and have enabled development of self-assembling peptide fibers, tapes, nanoparticles, and hydrogels.7-13 Furthermore, by connecting two or more peptide using short linkers, fairly complex assembled structures can be formed as result of the well-defined folding pattern of each single peptide folding motif.14-18 Peptides have also been conjugated to larger synthetic polymeric backbones, such as poly(ethylene glycol) (PEG), in order to create a wider range of supramolecular hybrid materials and nanostructures,19-21 including peptide-polymer hybrid hydrogels, fibrils, spherulities and micells.21-26 PEG is an attractive component in hydrogels and have been extensively investigated for 3D cell culture, and is used in a large number of clinical and pharmaceutical applications.27 By conjugating peptides to PEG, responsive hydrogels can be obtained as result of specific peptide-mediated self-assembly. These can in turn be proteolytically degraded, assembled or disassembled, and show other structural rearrangements as a result of changes in temperature, ionic strength or pH.

28-31

In addition,

peptides can provide highly defined intra- and intermolecular interactions. Highly defined interactions are crucial when developing supramolecular hydrogels with tunable mechanical and structural properties. For example, beta strand forming motifs are often used for selfassembly of peptide-based hydrogels, although the ability to control the gelation is often rather limited. In contrast, although less prone for gelation, α-helical coiled coil forming peptides exhibit much better control over the self-assembly process and the well-established design rules allows for highly defined interactions.32 However, limited progress has been made in exploiting the programmable interactions involved in coiled coil oligomerization and folding for controlling the rheological properties of hydrogels.

3 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

We have recently described a set of four 28 amino acid residue peptides that heterodimerize and fold into parallel coiled coils.33 The peptides are non-orthogonal and can form four different heterodimers that display large differences in affinities, which in turn promotes thermodynamic self-sorting. Here, we have conjugated these peptides to 4-armed star-shaped PEGs and investigated their assembly into supramolecular PEG-peptide networks and hydrogels (Figure 1). Four different PEG-peptide conjugates have been synthesized in order to address the hypothesis that mechanical properties of peptide-based hydrogels can be controlled using peptides with different affinities for dimerization. Additionally, we also explore peptide self-sorting using mutually complementary peptides to dynamically tune the self-assembly process of these nanomaterials. The proposed strategy offer means to control the interactions involved in self-assembly of peptide-polymer conjugates and consequently the possibilities to rationally tune dynamical and mechanical properties of the resulting materials.

4 ACS Paragon Plus Environment

Page 4 of 24

Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Figure 1. (A) Four different peptides are conjugated to maleimide functionalized 4-armed star-shaped PEGs via a Michael addition reaction. (B) The four different PEG-peptide conjugates can be combined in four different ways. Due to the difference in affinity for heterodimerization between peptides, the mechanical properties of the resulting PEG-peptide networks will be different.

Experimental Section General. Maleimide functionalized 4-armed star-shaped poly(ethylene glycol) (MW = 10 kDa, PDI =1.04) was purchased from JenKem Technology USA Inc., and is denoted PEG in this paper. PB denotes phosphate buffer (mono- and disodium phosphate, 10 mM, pH 7) unless otherwise stated. PBS denotes phosphate buffered saline (10 mM phosphate buffer, 2.7 mM KCl, 137 mM NaCl , pH 7.4). Peptide concentrations were determined by UV absorbance of tryptophan residues (ε280(Trp) = 5690 M-1 cm-1) in a UV-2450 from Shimadzu. PEG-peptide conjugates were always mixed in 1:1 ratios unless otherwise stated.

Peptides. EI-, EV-, KI- and KV peptides with a terminal Cys were ordered and customsynthesized by GL Biochem Ltd and used as received, whereas peptides without a terminal Cys were synthesized in house as described previously.33 Briefly, the in house synthesis was performed using a standard Fmoc protocol with HCTU as coupling agent. The peptides were purified by reverse phase HPLC and identified by MALDI-ToF spectroscopy with α-cyano-4hydroxycinnamic acid as matrix.

PEG-peptide conjugation. Cys containing peptides were dissolved in PBS and reduced in molar excess of dithiothreitol (DTT) for 30 min. DTT was subsequently removed by size exclusion on a PD MidiTrap G-10 from GE Healthcare according to manufacturer’s protocol. Reduced peptides were immediately eluted into a solution of PEGs (2:1 ratio of

5 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

peptide:maleimide). Samples were reacted overnight in PBS, purified by dialysis (MWCO 6-8 kDa, Spectra/Por 4 from Spectrum Laboratories Inc.) and finally freeze-dried to dry powder. 1

H-NMR was recorded on a Varian 300 spectrometer and used to verify the conjugation of

peptides by the disappearing maleimide (300 MHz, D2O, δ): 6.86 (s, 2H, C-H) and introduction of new peaks from peptide residues (e.g. Trp (300 MHz, D2O, δ): 7.3 (m, 4H, ArH). DCl was added to pKI4 and pKV4 whereas NaOD was added to pEI4 and pEV4, to increase solubility.

Circular Dichroism (CD) Spectroscopy. CD spectra were recorded using a ChirascanTM Spectropolarimeter from Applied Photophysics with a TC 125 temperature controller from Quantum Northwest. CD spectra were assessed from 195 to 250 nm at 0.5 nm steps. Thermal stability curves were measured by recording complete spectra in steps of 5 °C from 5 to 90 °C at a rate of roughly 50 °C per hour. The melting temperature (Tm) was estimated by calculating the derivative of each thermal stability curve. All data were converted to mean residue ellipticity (MRE). All samples were measured at 0.005 % (w/v), equivalent to a total peptide concentration of 10 µM, dissolved in PB.

Dynamic Light Scattering (DLS). DLS was measured on a ZetaSizer Nano ZS90 from Malvern Instruments Inc. All individual PEG-peptide conjugates were filtered through 0.2 µm low protein binding filters prior evaluation. Samples were mixed and measured over 4 h for each experiment. Temperature was kept at 20 °C and samples were measured at 0.125 % (w/v), equivalent to a total peptide concentration of 250 µM in PB.

6 ACS Paragon Plus Environment

Page 6 of 24

Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Turbidity measurements. The turbidity was measured by recording absorbance spectra with a UV-2450 from Shimadzu. A spectrum was acquired every hour for 12 hours and an aggregation index was calculated by the ratio of A340/A280.

Light Microscopy. Light micrographs were captured using a Nikon ECLIPSE Ti microscope at 20X using phase contrast enhancement. PEG-peptide conjugates were imaged at 1 % (w/v) in PB, equivalent to a total peptide concentration of 2 mM. Samples were incubated over night before analysis.

Scanning electron microscopy. Samples were imaged using a Leo Gemini from Zeiss with an acceleration voltage of 5 kV. The morphology of the PEG-peptide assemblies was fixated with 1 % glutaraldehyde over night at 4° C. Samples were sequentially chemically dried using increasing concentrations of ethanol in water (50, 70, 80, 90, 100 %) and finally 100 % hexamethyldisilazane for 15 min each. Samples were sputter coated with platinum prior to imaging.

Rheology. Rheology measurements were carried out on a MCR 102 rheometer from Anton Paar using a 25 mm 1° cone plate at 20 °C. Sample concentration was varied from 1 to 2.5 % (w/v). Gel properties were monitored via (1) frequency sweep measurements at fixed strain of 1 %, and (2) amplitude sweep measurements at fixed angular frequency of 5 rad/s. Time dependent gelling was assessed over 30 min at 1 % strain and 5 rad/s. Gel recovery was evaluated at 1 % strain and 5 rad/s for 15 min, after samples had been subjected to 1000 % strain at 6 rad/s for 180 s.

Results and Discussion

7 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Peptides and peptide conjugation to PEG. The peptides EI, EV, KI and KV (Table 1, Figure 2) are random coils as monomers and designed to heterodimerize and fold into parallel coiled coils at neutral pH.33 Complementary charged residues at the e and g position, either glutamic acid (Glu, E) or lysine (Lys, K), favor heterodimerization. In order to vary the affinity for dimerization EI and KI have mostly isoleucine (Ile, I) at position a, whereas EV and KV have valine (Val, V) at this position. The Asn (N) at position a in the third heptad promotes parallel alignment and prevents out-of-register structures. All peptides contain a tryptophan (Trp, W) residue as a chromophore to accurately determine the concentration of peptides in the sample. The peptides are promiscuous and EI, EV, KI and KV can be combined into four different heterodimers, EV/KV, EI/KV, EV/KI and EI/KI respectively. The different combinations show disassociation constants (Kd) spanning from the low micromolar to the picomolar range (Table S1). The large differences in affinities promotes thermodynamic self-sorting, which makes it possible to obtain defined stoichiometries of different heterodimers in complex mixtures of peptides and to dynamically exchange peptides in specific and already formed dimers.33

To enable peptide conjugation to the maleimide functionalized 4-arm-PEG, an off-heptad cysteine (Cys, C) residue was included at the peptide C-terminal in EI and EV and at the Nterminal in KI and KV. The formation of the thioether bond upon conjugation and complete absence of unreacted maleimides were verified by 1H-NMR (Figure S1). The resulting PEGpeptide conjugates were denoted pEV4, pEI4, pKV4 and pKI4.

Table 1. Peptide sequences with corresponding heptad register. Peptides for conjugation to 4-armed starshaped PEGs also included an off-heptad Cys residue at the N-terminal in KI and KV and at the Cterminal in EI and EV.

8 ACS Paragon Plus Environment

Page 8 of 24

Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Peptide sequences with corresponding heptad register Peptide

gabcdef

gabcdef

gabcdef

gabcdef

EV

EVSALEK

EVSALEK

ENSALEW

EVSALEK

KV

KVSALKE

KVSALKE

KNSALKW

KVSALKE

EI

EIAALEK

EIAALEK

ENAALEW

EIAALEK

KI

KIAALKE

KIAALKE

KNAALKW

KIAALKE

Figure 2. Helical wheel representation of the heterodimeric coiled coils.

Characterization of peptide secondary structure. It has previously been shown that coiled coils tend to retain their ability to fold when conjugated to linear poly(ethylene glycol).26,34 Pegylation of coiled coil peptides have even been showed to induce folding and to stabilize their secondary structure.34 Circular dichroism (CD) spectroscopy was used to investigate if conjugation of the peptides to the PEGs affected their ability to dimerize and fold. The peptide secondary structure of all four PEG-peptide conjugates, alone and when combined, was investigated (Figure 3, Table 2, Figure S2). The CD spectra of pEV4 and pKV4 showed a distinct minimum at 198 nm, indicating that the peptides are random coil at neutral pH (Figure 3A). In contrast, the CD spectra of pEI4 and pKI4 did show presence of α-helical secondary structure, but to varying degree. pEI4 showed only minor helicity with a melting temperature (Tm) well below 20 °C (Figure 3B). pKI4 showed a substantial helicity with distinct minima at 208 and 222 nm. The ratio of the mean residue ellipticity at 222 and 208 nm (MRE222/MRE208) was about 0.98 at 20 °C, where a ratio close to 1 or higher is indicative

9 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

of peptides folded into well-defined coiled coils.35 The estimated Tm was 43 °C, which is significantly higher than for the corresponding non-pegylated KI (Table S1). The increased helicity of the peptides in pKI4 compared to the other PEG-peptide conjugates is likely due to the beneficial packing of the Ile residues in the hydrophobic core in combination with the abundance of Lys residues at the dimer interface.36 The conformational flexibility of the Lys side-chain allows the ε-amine groups to rearrange to maximize their separation enough for peptides to overcome inter-peptide charge repulsion. This is less likely to occur for the shorter γ-carboxyls in Glu residues that are abundant in EI. EI is thus more effective in preventing homo-association compared to KI. In addition, since the non-pegylated peptides only show small to no tendencies of folding and forming homomeric species,33 it can be concluded that PEG conjugation induce folding of the peptides in pEI4 and pKI4. It is, however, not clear from this result alone if the peptides in a PEG-peptide conjugate fold individually or associate intramolecularly within the conjugate, or if the peptide homoassociate intermolecularly with peptides on another PEG-peptide conjugate, thus forming larger networks. Dynamic light scattering data (vide infra), however, showed that the latter was not the case.

10 ACS Paragon Plus Environment

Page 10 of 24

Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Figure 3. (A) CD spectra at 20 °C for individual PEG-peptide conjugates and (B) corresponding thermal denaturation curves. (C) CD spectra at 20 °C for combined complementary PEG-peptide conjugates and (D) corresponding thermal denaturation curves. Data points for curves in panel B and D can be found in the Supporting Information.

Table 2. MRE222, MRE222/ MRE208 and melting temperatures (Tm) of the individual and combinations of PEG-peptide conjugates at pH 7. MRE222 at 20 °C PEG-peptide

(103 deg cm2 dmol res-1)

MRE222/ MRE208 at 20 °C

Tm (°°C)

pEI4

-9.1

0.67

< 20

pEV4

-3.8

0.39

90

pEV4/pKV4

-14.3

0.76

51.3 ± 2.8

To further study the folding of the peptides in the individual PEG-peptide conjugates, the pH was varied from 4 to 10 and the changes in helicity and Tm of the peptides were examined (Figure 4, Table S2). The corresponding non-pegylated peptides do homoassociate and fold at pH

values

close

to their respective

isoelectric

point (pI).33 This is due

to

protonation/deprotonation of the amino acid residue side chains at e and g positions, rendering them uncharged and thus reducing the intermolecular charge repulsion. EI and EV (pI ~ 4) homoassociate at acidic pH whereas KI and KV (pI ~ 10.5) homoassociate at basic pH. As expected, pEI4 showed a transition from random coil to more α-helical when lowering the pH. However, pKI4 did not display any large increase in helicity when raising the pH from neutral to basic pH. A minor loss in helicity was seen at acidic pH but the peptides did not completely unfold within the measured pH range. The lack of unfolding of peptides in pKI4 at acidic pH is probably a result of the conformational flexibility of the Lys ε-amine group as discussed above. Furthermore, a clear difference in maximum helicity could be noted between pEI4 and pKI4 close to their respective pI. This difference can be explained by the number of hydrogen bonds that can be formed between the e and g amino acid residues in two homoassociated peptides. At acidic pH, each Glu-Glu’ interaction between two homoassociated EI results in formation of two hydrogen bonds. In contrast, two homoassociated KI at basic pH can only form one hydrogen bond for each Lys-Lys’ interaction due to steric hindrance, thus making it less stable compared to two homoassociated EI at acidic pH. Similar behavior as for pEI4 and pKI4 regarding folding at different pH has been described in other linear PEG-coiled coil systems with similar peptide sequences.26 Furthermore, a similar trend as for pEI4 could be seen for pEV4 with increasing helicity at acidic pH. In contrast, pKV4 showed no tendencies to

12 ACS Paragon Plus Environment

Page 13 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

fold regardless of pH. This shows that the better hydrophobic core packing provided by Ile as compared to Val in pKV4, is essential for pKI4 to be able to homoassociate.

Figure 4. The mean residue ellipticity at 222 nm at different pH, demonstrating the pH-induced homoassociation of each individual PEG-peptide conjugate at 20 °C.

When mixing the different complementary PEG-peptide conjugates, an increased helicity was seen for all combinations (Figure 3C). CD spectra at 20 °C of pEI4/pKI4 and pEI4/pKV4 showed MRE222/MRE208 ratios of 1.00 and 0.99, indicative of well-defined coiled coils. The other two combinations, pEV4/pKI4 and pEV4/pKV4, showed MRE222/MRE208 ratios that were slightly lower, 0.94 and 0.76, respectively at 20 °C. In the case of pEV4/pKV4, this is a result of the less favorable packing of the Val residues in the hydrophobic core compared to Ile. For pEV4/pKI4, on the other hand, the slightly lower ratio indicates that PEG conjugation of KI to a certain extent affects the association with EV in pEV4.

Thermal denaturation experiments further confirmed that all complementary PEG-peptide conjugates folded as a result of peptide-mediated intermolecular interactions (Figure 3D). The thermal stabilities of the individual PEG-peptide conjugates were significantly lower as compared to when mixed with a complementary PEG-peptide conjugate. The assembly of 13 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

pEI4/pKI4 showed the highest thermal stability followed by pEI4/pKV4, both with Tm > 90 °C. The Tm of pEV4/pKI4 was 73 °C and pEV4/pKV4 was the least stable with a Tm of 51 °C. Except for pEV4/pKI4, substantially higher Tm values were obtained for the interacting complementary PEG-peptide conjugates as compared to the corresponding folded nonpegylated coiled coils (Table S1). The higher stability is likely a combined effect of pegylation and positive cooperativity caused by formation of a supramolecular PEG-peptide network. The affinities and thermal stabilities of the non-pegylated coiled coils EI/KV and EV/KI are very similar, but interestingly the thermal stabilities of the corresponding PEGpeptide conjugates differ. Since peptides in pKI4 are already folded, and thus most likely homoassociated to some extent, interactions with the complementary peptides in pEV4 are energetically less favorable as compared to the corresponding interaction between pEI4 and pKV4. The thermal stability of pEV4/pKI4 is, however, higher than for the corresponding nonpegylated coiled coils and pKI4 alone, indicating that the peptides indeed heterodimerize to some extent when combined.

Self-assembly of PEG-peptide networks. The ability for the PEG-peptide conjugates to selfassemble into larger PEG-peptide networks was investigated using dynamic light scattering (DLS). DLS measurements were carried out at relatively low concentrations of the PEGpeptide conjugates (0.125% w/v, 250 µm total peptide concentration) to avoid formation of nonergodic hydrogels which would complicate the analysis. This concentration was, however, still high enough to readily promote and follow the self-assembly process of the PEG-peptide conjugates into larger supramolecular PEG-peptide networks. None of the individual PEGpeptide conjugates homoassociated intermolecularly into large assemblies at neutral pH as indicated by the single relaxation mode at short lag times (Figure 5). When mixing pEV4 and pKV4 the correlation function progressively shifted towards longer lag times with two modes 14 ACS Paragon Plus Environment

Page 14 of 24

Page 15 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

of relaxation (Figure 5D). This suggests a slowing of solution dynamics caused by the formation of larger assemblies with coexisting populations of smaller assemblies. Similar results has previously been reported for assemblies of peptides conjugated to N-(2hydroxypropyl)methacrylamide. 19

Figure 5. (A-C) Schematic representation and (D-K) DLS correlation functions showing the self-assembly and self-sorting of the PEG-peptide conjugates. (D-F) Mixing pEV4 and pKV4 induces self-assembly and formation

15 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

of a supramolecular PEG network, as seen by increased lag times. Addition of the non-pegylated peptide EI disassembles the PEG-peptide network due to formation and self-sorting into the more thermodynamically favored 4EI/pKV4 over pEV4/pKV4. Addition of non-pegylated KI induces further self-sorting and formation of the thermodynamically favored EI/KI over 4EI/pKV4. This makes pKV4 available to associate with pEV4 resulting in reassembly of the pEV4/pKV4 network. (G-K) Similar as for (D-F) but with other PEG-peptide conjugate combinations.

The dissociation constant (Kd) of the non-pegylated peptide heterodimers are Kd(EI/KI) < 0.1 nM, Kd(EI/KV) and Kd(EV/KI) ≈ 10 nM and Kd(EV/KV) ≈ 1 µM. When all four nonpegylated peptides are combined, the highest and lowest affinity dimers (EI/KI and EV/KV) form at the cost of those with intermediate affinities (EI/KV and EV/KI) because of the large differences in affinities for heterodimerization.33 Interestingly, this ability of the peptides to self-sort was retained when conjugated to PEG. This is indicated by the almost instant disassembly of the pEV4/pKV4 networks upon addition of non-pegylated EI (125 µM), seen as a change from two to one mode of relaxation with lag times comparable to the individual PEG-peptide conjugates (Figure 5E). The non-pegylated EI thus penetrates the PEG-peptide network and scavenge pKV4 to form the more thermodynamically favored 4EI/pKV4. Subsequent addition of non-pegylated KI (125 µM) lead to formation of EI/KI heterodimers, allowing reassembly of the pEV4/pKV4 network (Figure 5F). The same result was seen when adding non-pegylated KI prior to non-pegylated EI (Figure S3). The disassembly of the initial network by non-pegylated KI was not as rapid as with EI, probably due to the slightly higher stability of KI homomers.

Self-assembly into PEG-peptide networks and self-sorting were seen for both pEV4/pKI4 and pEI4/pKV4. For pEV4/pKI4 there was an instant increase in lag times with a progression from one to two modes of relaxation (Figure 5G). Addition of non-pegylated EI instantly 16 ACS Paragon Plus Environment

Page 16 of 24

Page 17 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

disassembled the pEV4/pKI4 network (Figure 5H). pEI4/pKV4 did also show a progressive growth of a PEG-peptide network, which could be disassembled by addition of non-pegylated KI (Figure 5I-J). However, pEI4/pKV4 did only show one mode of relaxation as the network was formed. This was also the case for pEI4/pKI4 (Figure 5K). The absence of larger species, as indicated by the single mode of relaxation at relatively short lag times, for these combinations is likely due to assembly of structures that precipitated during the time course of the measurements and thus were too large to be detected by DLS. To verify this, the turbidity changes were monitored over time by measuring the absorbance of the samples. An aggregation index was calculated as the ratio of the absorption at 340 and 280 nm (Figure S4). pEV4/pKV4 and pEV4/pKI4 did not show any major changes in turbidity over time, indicating presence of only relatively small assemblies. However, both pEI4/pKI4 and pEI4/pKV4 showed major increase in aggregation index over time, indicating formation of larger assemblies compared to pEV4/pKV4 and pEV4/pKI4. Observing pEI4/pKI4 and pEI4/pKV4 under a conventional light microscope confirmed that larger microscopic assemblies were formed (Figure 6A-B). As indicated by DLS, the pEI4/pKV4 network could also be disassembled by addition of non-pegylated KI (Figure 6C). No visible microscopic assemblies could be seen for pEV4/pKV4 nor pEV4/pKI4. Furthermore, scanning electron micrographs of the assemblies revealed structures with a fibrilliar and porous morphology, similar to those seen in many other helical peptide-based gelation systems.16,37

17 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. (A) Light microscopy image of pEI4/pKI4. (B) Light microscopy image of pEI4/pKV4 and (C) after addition of KI to pEI4/pKV4. (D) Scanning electron micrographs of pEI4/pKI4, and (E-F) pEI4/pKV4.

Formation of hydrogels. Since pEI4/pKI4 and pEI4/pKV4 did assemble into microscopic PEG-peptide networks already at fairly low concentrations (0.125 % w/v), it was possible to induce formation of hydrogels by increasing the concentration above 1 % (w/v). The pEV4/pKV4 and pEV4/pKI4 did not assemble into hydrogels at these concentrations. Tabletop rheology showed that hydrogels comprised of pEI4/pKI4 were markedly more rigid than hydrogels of pEI4/pKV4 at the same concentration (Figure 7A-C).

Bulk oscillatory rheology was employed to further characterize the viscoelastic properties of the pEI4/pKI4 and pEI4/pKV4 hydrogels. Gel strengths and viscoelastic behavior were assessed by frequency sweep measurements and confirmed the observations from the table-top rheology (Figure 7D-E). Hydrogels formed from pEI4/pKV4 showed storage moduli (G’) of about 200 Pa whereas G’ for pEI4/pKI4 hydrogels was about a factor five higher, 1000 Pa, at 2.5 % (w/v). This G’ interval spans the range of most coiled coil-based hydrogels previously reported.8,31,38 Furthermore, pEI4/pKI4 showed a typical solid gel-like behavior (G’ > G’’), 18 ACS Paragon Plus Environment

Page 18 of 24

Page 19 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

whereas pEI4/pKV4 instead showed a liquid-like behavior (G’ < G’’) at low frequencies with a cross point at 0.5 rad/s. Both gels exhibited, as expected, a decrease in gel strength with decreasing PEG-peptide conjugate concentrations. Furthermore, the hydrogels were formed almost instantaneous since more than 50 % of the final G’ was reached within 15 seconds after mixing the two PEG-peptide conjugates (Figure 7F-G). This is more rapid than for example the beta-hairpin MAX1 hydrogel at comparable concentrations.6 One hallmark of supramolecular hydrogels is shear thinning behavior where the viscosity decrease with increasing strain rate due to the disruption of supramolecular cross-links by the applied strain.39,40 However, they can typically also quickly recover their viscoelastic properties when the strain is discontinued since the supramolecular interactions can reform. This is not seen in chemically crosslinked polymeric networks.9,11 As expected, both hydrogels showed a decrease in viscosity as more strain was applied (Figure S5A-B). Furthermore, both hydrogels also showed quick recoveries at concentrations of 2.5 and 2.0 % (w/v) (Figure S5C-D), whereas at lower concentrations the hydrogels did not fully recover during the time course of the measurement.

19 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. (A–C) Table-top rheology showing the difference in gel strength of pEI4/pKI4 and pEI4/pKV4, respectively. (D) and (E) shows amplitude sweep data for different gel concentration of (D) pEI4/pKI4 and (E) pEI4/pKV4. (F) and (G) shows the formation and the gel strength of (F) pEI4/pKI4 and (G) pEI4/pKV4 gels at 2.5 % (w/v) as a function of time.

Conclusions. In this work, we have investigated the possibility to modulate the self-assembly process of polymeric networks and hydrogels, using four different heterodimeric coiled coil peptides with different, but defined, affinities for dimerization. The peptides retained their ability to dimerize and fold into coiled coils when conjugated to 4-armed star-shaped PEGs. Depending on the combinations of peptides used, varying assembly kinetics and temperature 20 ACS Paragon Plus Environment

Page 20 of 24

Page 21 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

stabilities were observed. Peptides with high affinities for dimerization resulted in rapid assembly of large supramolecular networks with high thermal stability (Tm > 90 °C), whereas peptides with lower affinities resulted in smaller and less stable assemblies. Hydrogels were obtained at concentrations above 1 % (w/v) for two combinations of PEG-peptide conjugates. The self-assembly into PEG-peptide networks were rapid and commenced immediately when complementary PEG-peptide conjugates were mixed, and shear thinning- and self-healing hydrogels were formed within seconds to minutes. The storage modulus of the hydrogels was about 200 Pa when using coiled coils with nM affinities, and about a factor five higher, 1000 Pa, when using peptides with pM affinities. In addition to the possibility to generate hydrogels with different rheological properties, the large differences in affinities for peptide dimerization did enable dynamic modulation of the self-assembly process by means of peptide self-sorting. By sequential additions of non-pegylated peptides with increasing affinities for dimerization it was possible to disassemble and reassemble the supramolecular PEG-peptide networks. Hence, the possibility to combine multiple complementary coiled coils opens up for numerous strategies to tailor hydrogel properties and to optimize their performance for various biomedical applications.

Associated content Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: ############ Additional tables, 1H-NMR of PEG-peptide conjugates, CD-data and thermal stability curves of individual and mixed PEG-peptide conjugates at different temperatures and pH, additional

21 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

DLS-data, turbidity measurements, additional rheology data, purity and identity of nonpegylated EI and KI.

Author Information Corresponding Author *[email protected] Notes The authors declare no competing financial interest

Acknowledgements This work was supported by the Swedish Research Council (621-2011-4319), the Swedish Foundation for Strategic Research (ICA10-0002), Linköping University and the Swedish Government Strategic Research Area in Materials Science on Functional Materials at Linköping University (Faculty Grant SFO-Mat-LiU No 2009 00971). During this study C.A. was enrolled in the graduate school Forum Scientium. Mats Sandberg and Anurak Sawatdee at Acreo Swedish ICT (Norrköping) are acknowledged for assistance with the rheometer equipment.

22 ACS Paragon Plus Environment

Page 22 of 24

Page 23 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

References

(1) Seliktar, D. Science 2012, 336, 1124-1128. (2) Murphy, S. V.; Atala, A. Nat. Biotech. 2014, 32, 773-785. (3) Rape, A. D.; Zibinsky, M.; Murthy, N.; Kumar, S. Nat. Commun. 2015, 6. (4) Goktas, M.; Cinar, G.; Orujalipoor, I.; Ide, S.; Tekinay, A. B.; Guler, M. O. Biomacromolecules 2015, 16, 1247-1258. (5) Paul, A.; Manoharan, V.; Krafft, D.; Assmann, A.; Uquillas, J. A.; Shin, S. R.; Hasan, A.; Hussain, M. A.; Memic, A.; Gaharwar, A. K.; Khademhosseini, A. J. Mat. Chem. B 2016, 4, 3544-3554. (6) Stevens, M. M.; George, J. H. Science 2005, 310, 1135-1138. (7) Reches, M.; Gazit, E. Science 2003, 300, 625-627. (8) Banwell, E. F.; Abelardo, E. S.; Adams, D. J.; Birchall, M. A.; Corrigan, A.; Donald, A. M.; Kirkland, M.; Serpell, L. C.; Butler, M. F.; Woolfson, D. N. Nat. Mater. 2009, 8, 596600. (9) Schneider, J. P.; Pochan, D. J.; Ozbas, B.; Rajagopal, K.; Pakstis, L.; Kretsinger, J. J. Am. Chem. Soc. 2002, 124, 15030-15037. (10) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Proc. Natl. Acad. Sci. USA 2002, 99, 51335138. (11) Aggeli, A.; Bell, M.; Boden, N.; Keen, J. N.; Knowles, P. F.; McLeish, T. C.; Pitkeathly, M.; Radford, S. E. Nature 1997, 386, 259-262. (12) Pandya, M. J.; Spooner, G. M.; Sunde, M.; Thorpe, J. R.; Rodger, A.; Woolfson, D. N. Biochemistry 2000, 39, 8728-8734. (13) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Science 2001, 294, 1684-1688. (14) Zhou, M.; Bentley, D.; Ghosh, I. J. Am. Chem. Soc. 2004, 126, 734-735. (15) Staples, J. K.; Oshaben, K. M.; Horne, W. S. Chem. Sci. 2012, 3, 3387-3392. (16) Faruqui, N.; Bella, A.; Ravi, J.; Ray, S.; Lamarre, B.; Ryadnov, M. G. J. Am. Chem. Soc. 2014, 136, 7889-7898. (17) Gradišar, H.; Božič, S.; Doles, T.; Vengust, D.; Hafner-Bratkovič, I.; Mertelj, A.; Webb, B.; Šali, A.; Klavžar, S.; Jerala, R. Nat. Chem. Biol. 2013, 9, 362-366. (18) Aili, D.; Tai, F.-I.; Enander, K.; Baltzer, L.; Liedberg, B. Angew. Chem. Int. Ed. 2008, 47, 5554-5556. (19) Yang, J.; Wu, K.; Konak, C.; Kopecek, J. Biomacromolecules 2008, 9, 510-517. (20) Yang, J.; Xu, C.; Wang, C.; Kopecek, J. Biomacromolecules 2006, 7, 1187-1195. (21) Marsden, H. R.; Handgraaf, J.-W.; Nudelman, F.; Sommerdijk, N. A. J. M.; Kros, A. J. Am. Chem. Soc. 2010, 132, 2370-2377. (22) Hamley, I. W. Biomacromolecules 2014, 15, 1543-1559. (23) Burkoth, T. S.; Benzinger, T. L. S.; Urban, V.; Lynn, D. G.; Meredith, S. C.; Thiyagarajan, P. J. Am. Chem. Soc. 1999, 121, 7429-7430. (24) Castelletto, V.; McKendrick, J. E.; Hamley, I. W.; Olsson, U.; Cenker, C. Langmuir 2010, 26, 11624-11627. (25) Pechar, M.; Kopečková, P.; Joss, L.; Kopeček, J. Macromol. Biosci. 2002, 2, 199-206. (26) Vandermeulen, G. W. M.; Tziatzios, C.; Duncan, R.; Klok, H.-A. Macromolecules 2005, 38, 761-769. (27) Roberts, M. J.; Bentley, M. D.; Harris, J. M. Adv. Drug. Deliv. Rev. 2002, 54, 459-476. (28) Rubert Perez, C. M.; Rank, L. A.; Chmielewski, J. Chem. Commun. 2014, 50, 81748176.

23 ACS Paragon Plus Environment

Biomacromolecules

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(29) Stahl, P. J.; Romano, N. H.; Wirtz, D.; Yu, S. M. Biomacromolecules 2010, 11, 23362344. (30) Xu, C.; Breedveld, V.; Kopecek, J. Biomacromolecules 2005, 6, 1739-1749. (31) Petka, W. A.; Harden, J. L.; McGrath, K. P.; Wirtz, D.; Tirrell, D. A. Science 1998, 281, 389-392. (32) Woolfson, D. N. Woolfson, D. N. Adv. Prot. Chem. 2005, 70, 79-112. (33) Aronsson, C.; Dånmark, S.; Zhou, F.; Öberg, P.; Enander, K.; Su, H.; Aili, D. Sci. Rep. 2015, 5, 14063. (34) Vandermeulen, G. W. M.; Tziatzios, C.; Klok, H.-A. Macromolecules 2003, 36, 41074114. (35) Lau, S. Y.; Taneja, A. K.; Hodges, R. S. J. Biol. Chem. 1984, 259, 13253-13261. (36) Zhu, B. Y.; Zhou, N. E.; Kay, C. M.; Hodges, R. S. Prot. Sci. 1993, 2, 383-394. (37) Mondal, S.; Adler-Abramovich, L.; Lampel, A.; Bram, Y.; Lipstman, S.; Gazit, E. Nat. Commun. 2015, 6. (38) Nowak, A. P.; Breedveld, V.; Pakstis, L.; Ozbas, B.; Pine, D. J.; Pochan, D.; Deming, T. J. Nature 2002, 417, 424-428. (39) Owen, D. H.; Peters, J. J.; Katz, D. F. Contraception 2000, 62, 321-326. (40) Prud'homme, R. K.; Wu, G.; Schneider, D. K. Langmuir 1996, 12, 4651-4659. Insert Table of Contents artwork here

24 ACS Paragon Plus Environment

Page 24 of 24