Tuning the pH Responsiveness of β-Hairpin Peptide Folding, Self

Aug 7, 2009 - Karthikan Rajagopal,† Matthew S. Lamm,‡ Lisa A. Haines-Butterick,† Darrin J. Pochan,*,‡ and Joel P. Schneider*,†. Department o...
0 downloads 0 Views 4MB Size
Biomacromolecules 2009, 10, 2619–2625

2619

Tuning the pH Responsiveness of β-Hairpin Peptide Folding, Self-Assembly, and Hydrogel Material Formation Karthikan Rajagopal,† Matthew S. Lamm,‡ Lisa A. Haines-Butterick,† Darrin J. Pochan,*,‡ and Joel P. Schneider*,† Department of Chemistry and Biochemistry, Department of Materials Science and Engineering, and Delaware Biotechnology Institute, University of Delaware, Newark Delaware 19716 Received May 13, 2009; Revised Manuscript Received June 29, 2009

A design strategy to control the thermally triggered folding, self-assembly, and subsequent hydrogelation of amphiphilic β-hairpin peptides in a pH-dependent manner is presented. Point substitutions of the lysine residues of the self-assembling peptide MAX1 were made to alter the net charge of the peptide. In turn, the electrostatic nature of the peptide directly influences the solution pH at which thermally triggered hydrogelation is permitted. CD spectroscopy and oscillatory rheology show that peptides of lower net positive charge are capable of folding and assembling into hydrogel material at lower values of pH at a given temperature. The pH sensitive folding and assembling behavior is not only dependent on the net peptide charge, but also on the exact position of substitution within the peptide sequence. TEM shows that these peptides self-assemble into hydrogels that are composed of well-defined fibrils with nonlaminated morphologies. TEM also indicates that fibril morphology is not influenced by making these sequence changes on the hydrophilic face of the hairpins. Rheology shows that the ultimate mechanical rigidity of these peptide hydrogels is dependent on the rate of folding and self-assembly. Peptides that fold and assemble faster afford more rigid gels. Ultimately, this design strategy yielded a peptide MAX1(K15E) that is capable of undergoing thermally triggered hydrogelation at physiological buffer conditions (pH 7.4, 150 NaCl, 37 °C).

Introduction Wide ranging applications for hydrogels in biotechnology, such as scaffolds for tissue engineering,1 vehicles for controlled drug release,2 or actuators in microfluidic systems3 stems from their inherent physicochemical characteristics. These include a porous and elastic network structure, encapsulation of large amounts of water, responsive phase and volume transition behavior, favorable cell adhesion properties, biocompatibility, and controlled biodegradability.4-9 The utility of hydrogels for such applications can be further enhanced if their formation, as well as some of the above-mentioned characteristics, can be controlled in a stimulus-responsive manner. Toward this objective, a number of synthetic and natural polymers have been developed that form hydrogels in response to a specific stimulus such as pH,10-12 heat,13-16 or light.17-19 More recently, molecular self-assembly of designed peptides has emerged as a novel route to form responsive hydrogels.20-25 Peptide hydrogels have shown promise for use as scaffolds for tissue regeneration26,27 and release of drugs and growth factors.28,29 In this article we present a peptide design strategy that can be utilized to control the thermally triggered hydrogelation of β-hairpin peptides in a pH-responsive manner. The construction of peptide-based, self-assembled materials relies heavily on secondary structural building blocks, namely, R-helices30-33 and β-sheets.24,25,34 For example, the β-sheet motif has been extensively used in the preparation of a variety of supramolecular structures such as nanotubes,35 monolayers,36 * To whom correspondence should be addressed. Phone: 302-831-3024 (J.P.S.); 302-831-3569 (D.J.P.). Fax: 302-831-6335 (J.P.S.). E-mail: [email protected] (J.P.S.); [email protected]. † Department of Chemistry and Biochemistry. ‡ Department of Materials Science and Engineering and Delaware Biotechnology Institute.

fibrils,37 ribbons, and tapes.38 For the preparation of hydrogel materials, in particular, self-assembling β-strands that form fibrils rich in β-sheet structure has been well exploited.25,39-43 We reported earlier the design and characterization of a 20 amino acid peptide (MAX 1) that is capable of undergoing triggered hydrogelation.44,45 The primary sequence of MAX1 contains 20 amino acids (Figure 1), eight of which are lysine residues and nine of which are valine residues. When dissolved in aqueous solutions of low ionic strength at pH 9 and temperatures below ∼20 °C, the peptide remains unfolded. Mainly this is due to the fact that a large population of protonated lysine side chains would have to occupy one face of the molecule if the peptide were to fold. Due to electrostatics, this is an energetically unfavorable scenario. However, increasing the temperature favors the desolvation of the hydrophobic valines and leads to peptide folding and self-assembly, Figure 1. When assembled into the hydrogel network, the valine side chains are shielded from water. In this folding and self-assembly mechanism, the free energy gained in shielding the hydrophobic portion of MAX1 from water outweighs the electrostatic repulsions that must be overcome for the peptide to fold. In the folded state, MAX1 adopts a β-hairpin that contains two β-strands composed of alternating valine and lysine residues connected by a four residue type II′ turn sequence (-VDPPT-). The hairpin is amphiphilic in that its bottom face is composed of hydrophobic valine residues and the top face is composed of lysine residues. Small angle neutron scattering (SANS) and transmission electron microscopy (TEM) data show that the hairpin self-assembles laterally by forming a network of intermolecular hydrogen bonds that define the long axis of a given fibril; all the β-strands of the assembled hairpins are in register, affording fibrils of distinct diameter (∼3 nm).44,45 MAX1 also self-assembles in a facial manner by burying its

10.1021/bm900544e CCC: $40.75  2009 American Chemical Society Published on Web 08/07/2009

2620

Biomacromolecules, Vol. 10, No. 9, 2009

Rajagopal et al.

Figure 1. Folding and self-assembly mechanism of MAX1 leading to hydrogel formation. Increasing the temperature of a solution of unfolded peptide at pH 9 results in the formation of an amphiphilic β-hairpin structure that subsequently undergoes lateral and facial self-assembly to form a fibrillar supramolecular structure. Non-covalent cross-linking of the fibrils results in the formation of a macroscopic hydrogel. The sequence of MAX1 is shown below.

valines to form a bilayer that defines the thickness of a given fibril. Along the long axis of a given fibril, bilayer formation occurs in a regular fashion with one hairpin docked, and in register, with its partner. This arrangement shields the maximal amount of valine side chain surface area from water. However, imperfections in this mechanism occur where the face of one hairpin is rotated relative to its partner in the bilayer. This results in a site for nascent fibril growth in a new three-dimensional direction and constitutes the formation of an interfibril crosslink (Figure 1). These interfibril cross-links, in addition to fibril entanglements, contribute to the mechanical rigidity of the gel.46 Although, peptide folding and self-assembly events are discussed independently, their equilibria may be linked. For example, during gelation, unfolded peptides may fold at the surfaces of growing fibrils. In general, the kinetic rates of peptide folding and self-assembly dictates the number of cross-links that are formed during the gelation process and the number of crosslinks directly influences the mechanical rigidity of the gel. Slow kinetics of assembly leads to less rigid gels with fewer crosslinks. Fast kinetics lead to a more rigid gel containing more cross-links.47,48 Herein, we show that the temperature-induced folding and assembly of MAX1 is sensitive to pH. Importantly, this discovery facilitated the generation of a suite of designed peptides that undergo temperature-dependent folding, selfassembly, and hydrogelation in environments of distinct pH. Lastly, we show that hydrogels of varying mechanical rigidity can be produced at a given pH by altering the net charge of the peptide, which directly influences its rate of hydrogelation.

Experimental Section Peptide Synthesis. All the peptides were synthesized on a Rink amide resin via automated Fmoc peptide synthesis employing an ABI 433A peptide synthesizer and HCTU activation. The resulting dry resinbound peptides were cleaved and side-chain deprotected using a cocktail of TFA, thioanisole, ethanedithiol, and anisole in the ratio 90:5:3:2. After filtration and precipitation from cold ether, the crude peptide was purified by RP-HPLC using a preparative Vydac C18 peptide/protein column. The following gradient was employed: isocratic at 0% B for 2 min, then a linear gradient to 12% B over 3 min and then a linear gradient from 12% to 100% B over 176 min. Solvent A is 0.1% TFA in water and solvent B is 90% acetonitrile, 10% water, and 0.1% TFA. Identity of all the peptides was established by ESI mass spectrometry (see Supporting Information). Hydrogel Preparation. Bulk hydrogels can be formed as follows. A total of 2.0 mg of the appropriate peptide is placed in a 2 mL glass

vial. Milli Q water (200 µL) is added to dissolve the peptide and yield a clear solution. To this, 200 µL of appropriate buffer stock solution (100 mM buffer and 20 mM NaCl) is added, resulting in a 0.5 wt % peptide solution. After gently mixing, the solution is allowed to stand at a desired temperature until gelation is complete. Hydrogels formed in this manner are visually clear and self-supporting. Circular Dichroism Spectroscopy. Circular dichroism (CD) spectra were collected on a Jasco J-810 spectropolarimeter. pH- and temperature-dependent measurements for each of the peptides at 150 µM concentration were collected as follows. A 300 µM stock solution of a given peptide was prepared in water and kept on ice until used. A stock buffer (50 mM buffer and 20 mM NaCl) solution is also kept on ice. The following buffers were used for the following pH ranges: pH 6 (MES), pH 7 and 8 (BTP), and pH 9 and 9.7 (borate). Prior to the start of each experiment, an equal volume (∼125 µL) of the peptide stock and the appropriate buffer stock were gently mixed in a 1 mm quartz cell by cell inversion to yield the final peptide solution (150 µM peptide, 50 mM buffer, 10 mM NaCl). The cell was immediately placed in the cell holder pre-equilibrated at 5 °C. Temperaturedependent wavelength scans from 260 to 190 nm were collected every 5 °C from 5 to 60 °C, with a 10 min equilibration time at each temperature interval. The data in Figures 3-5 were extracted from the temperature-dependent wavelength spectra collected above. Kinetic experiments at 0.5 wt % peptide concentration were performed as follows. Peptide samples at 1 wt % concentration were prepared in water and mixed with buffer stock solutions (100 mM buffer and 20 mM NaCl) at room temperature. The contents were gently mixed and immediately transferred to a 0.1 mm path length cell pre-equilibrated to 50 °C. Time-dependent CD measurements at 216 nm were collected every 20 s for 1 h; data at every 100 s is shown for clarity. Concentrations of peptide solutions were determined by absorbance at 220 nm (ε ) 15750 cm-1 M-1). Mean residue ellipticity [θ] was calculated from the equation [θ] ) (θobs/10lc/r, where θobs is the measured ellipticity in millidegrees, l is the path length of the cell (centimeters), c is the concentration (molar), and r is the number of residues in the sequence. Error in the CD measurements (Figures 2-5) is less than 5% and originates from the error in sample concentration determination, which was determined by UV as discussed above. Oscillatory Rheology. Dynamic time, frequency, and strain sweep rheology experiments were performed on AR2000 Advanced Rheometer (TA Instruments) with 25 mm diameter parallel plate geometry at 50 °C. A 0.5 wt % peptide sample is prepared by mixing equal volumes of peptide stock (1 wt %) with an appropriate buffer solution pre-equilibrated in an ice bath and transferred immediately to the rheometer at 5 °C. The temperature is ramped from 5 to 50 °C in 100 s. The dynamic time sweep experiment, which measures the evolution of storage (G′) and loss modulus (G′′) with time, was performed for

pH Responsiveness of β-Hairpin Peptide Folding

Figure 2. Temperature-dependent formation of β-sheet structure is shown for 150 µM MAX1 at pH 8, 9, and 9.7. β-Sheet formation is monitored by measuring the mean residue ellipticity at 216 nm ([θ]216).

1 h at constant strain (0.2%) and frequency (6 rad/s). The time sweep data in Figure 6b was collected at a frequency of 6 rad/s using a strain of 0.2%. Independent dynamic frequency sweep (0.1-100 rad/s at 0.2% strain) and strain sweep (0.1-100% strain at 6 rad/s) experiments confirm that the time sweep measurements were performed within the linear viscoelastic regime. Transmission Electron Microscopy. Dilute samples of the peptides at 0.2% (w/v) were prepared in appropriate buffer (50 buffer, 10 mM NaCl) and equilibrated at 50 °C for 10 min. Small amounts (∼5 µL) of each hydrogel solution were applied to separate carbon-coated copper grids. Excess solution was removed by filter paper after 1 min. The grid was then washed with deionized water (2 × 100 µL). Then, a 1% (w/v) uranyl acetate aqueous solution was placed on each grid for negative staining and excess staining solution was blotted with filter paper and left to air-dry. Bright field images of the samples were taken on a JEOL 2000-FX transmission electron microscope at 200 kV accelerating voltage with a Gatan CCD camera.

Results and Discussion pH Sensitivity of Folding and Self-Assembly. When MAX1 is dissolved in pH 9 buffer of low ionic strength, it is capable of undergoing temperature-induced folding and self-assembly. Figure 2 shows CD data that monitors the mean residue ellipticity at 216 nm, an indicator of β-sheet structure, as a function of temperature. Spectra were obtained at a peptide concentration of 150 µM. At this low concentration there is not enough peptide to gel the buffer solution, but the peptide still folds and assembles enabling facile spectroscopic measurements of secondary structure formation. At pH 9.0 and at low

Biomacromolecules, Vol. 10, No. 9, 2009

2621

Figure 4. Fraction of β-sheet formed as a function of pH is shown for hairpins that vary in net positive charge according to Table 2. Spectra of peptides (150 µM) were collected at 50 °C.

Figure 5. Folding and self-assembly of V16E and K15E variants of MAX1 is monitored by measuring [θ]216 as a function of pH at 50 °C. The peptide concentration is 150 µM.

temperatures, MAX1 is unfolded, but as the temperature is increased beyond ∼25 °C, the hydrophobic effect, which is temperature-dependent,49-52 triggers folding and assembly. At pH 9, most of the lysine side chains are protonated, however, a small fraction are deprotonated53 and the peptide is poised to fold when sufficient thermal energy is supplied. The data in Figure 2 also show that this temperature-induced transition is sensitive to pH. At pH 9.7, a greater percentage of lysine side chains are deprotonated and the peptide folds and selfassembles at a lower temperature; it take less thermal energy to drive the hydrophobic collapse of the peptide. In contrast, at pH 8, the N-terminal amine as well as all of the lysine residues of the peptide are most likely fully protonated. As a result, the peptide is incapable of folding irrespective of

Figure 3. (a) The folding and self-assembly of MAX1 and MAX1(K19E) is monitored by measuring [θ]216 as a function of pH at 50 °C. (b) Folding and self-assembly is monitored under identical conditions for eight variants of MAX1, where each lysine of the peptide is replaced by a glutamic acid.

2622

Biomacromolecules, Vol. 10, No. 9, 2009

Rajagopal et al.

Figure 6. (a) Rates of β-sheet formation for peptides varying in net charge (Table 2) is monitored by measuring [θ]216 as a function of time at pH 8, 50 °C, and 0.5 wt % peptide. (b) Dynamic time sweep rheology showing the evolution of storage moduli with time at pH 8 and 50 °C for 0.5 wt % peptide. Table 1. Glutamic Acid Variants of MAX1 modification MAX1 MAX1 MAX1 MAX1 MAX1 MAX1 MAX1 MAX1 MAX1

D

(K19E) (K2E) (K4E) (K6E) (K8E) (K13E) (K15E) (K17E)

Table 2. MAX1 Variants with Differing Net Positive Charge

peptide sequence VKVKVKVKV PPTKVKVKVKV-NH2 VKVKVKVKVDPPTKVKVKVEV-NH2 VEVKVKVKVDPPTKVKVKVKV-NH2 VKVEVKVKVDPPTKVKVKVKV-NH2 VKVKVEVKVDPPTKVKVKVKV-NH2 VKVKVKVEVDPPTKVKVKVKV-NH2 VKVKVKVKVDPPTEVKVKVKV-NH2 VKVKVKVKVDPPTKVEVKVKV-NH2 VKVKVKVKVDPPTKVKVEVKV-NH2

temperature. This data indicates that the temperature-induced folding and self-assembly events leading to hydrogelation are sensitive to the net charge of the peptide, which is largely defined by the protonation states of the lysine residues occupying the hydrophilic face of the hairpin. This observation suggests that peptides can be designed to undergo thermally induced gelation at distinct values of pH by making point substitutions of the lysine residues, which change the net charge of the peptide. This hypothesis was first tested by replacing the C-terminal lysine of MAX1 with glutamic acid (side chain pKa ∼ 4.3), resulting in the peptide MAX1(K19E), Table 1. The ramification of making this change is best realized by comparing the charge state of both peptides at pH 7. MAX1 has a net charge of +9 due to the eight positively charged lysine side chains and the N-terminal amine. MAX1(K19E) has a net charge of +7. One would predict that, at a common pH, MAX1(K19E) should fold and assemble at a lower temperature compared to MAX1. This is because, for MAX1(K19E), less thermal energy is needed to initiate the hydrophobic effect, which must ultimately overcome the electrostatic repulsions centered on the hydrophilic face of the hairpin. Alternatively, at a common temperature, MAX1(K19E) should be capable of folding and assembling in environments of lower pH compared to MAX1. Figure 3a shows [θ]216 as a function of pH for MAX1 and MAX1(K19E) at a common temperature of 50 °C. This temperature provides sufficient thermal energy to trigger MAX1 folding at pH 9 and allows the pH-dependent folding behavior of peptides that derive from MAX1 to be comparatively evaluated. Low ionic strength buffers containing 10 mM NaCl were used for these studies to maximize the effects of pH on folding and assembly. The data clearly shows that folding can be triggered at lower values of pH by simply reducing the peptide’s electropositive charge. MAX1(K19E) is capable of folding and assembling to its full potential at pH 8, whereas MAX1 remains completely unfolded.

peptide

sequence

net charge at pH 7

MAX1 MAX1 (K15R) positive control MAX1 (K15T) MAX1 (K15I) MAX1 (K15E) MAX1 (K4E, K15E) MAX1 (V16E) alteration on the hydrophobic face

VKVKVKVKVDPPTKVKVKVKV-NH2 VKVKVKVKVDPPTKVRVKVKV-NH2

+9 +9

VKVKVKVKVDPPTKVTVKVKV-NH2 VKVKVKVKVDPPTKVIVKVKV-NH2 VKVKVKVKVDPPTKVEVKVKV-NH2 VKVEVKVKVDPPTKVEVKVKV-NH2 VKVKVKVKVDPPTKVKEKVKV-NH2

+8 +8 +7 +5 +8

The dependence of glutamic acid substitution with respect to its location on the hydrophilic face of the hairpin was investigated by sequentially replacing each lysine of MAX1 with a glutamic acid, Table 1. Figure 3b shows the folding propensity of each peptide variant as a function of pH at 50 °C. All of the peptides are capable of folding and assembling at lower values of pH than MAX1. When point substitutions are made at positions 8 or 13, the smallest effect is realized. However, when the lysine at position 4 is replaced with a glutamic acid, the resulting peptide (MAX1(K4E)) is capable of significant β-sheet formation even at pH 7, nearly two pH units lower than that required for MAX1. The remaining peptides show intermediate behavior. For example, MAX1(K2E) shows a mean residue ellipticity value consistent with the peptide in thermodynamic equilibrium between unfolded and the folded/self-assembled state. This data demonstrates that the exact location of the substitution influences the pH sensitivity of the temperatureinduced folding and assembly event. Substitutions made within the folded hairpin that alleviate the most dense regions of positive charge have the largest effect on the pH responsiveness. For example, when the lysines at positions 4 or 17 are replaced with glutamic acid, folding and assembly is more facile. All of the sequence variants of MAX1 listed in Table 1 carry a net charge of +7 at neutral pH. Additional peptides were synthesized that systematically vary in their charge state to determine if the pH sensitivity of the thermally induce folding event is sensitive to differences in the net electrostatics of the peptide. The lysine at position 15 was substituted with residues bearing positive (Arg), negative (Glu), and neutral (Thr, Ile) side chains to actively vary the net charge of the hairpin, Table 2. Position 15 was chosen for the substitutions based on the data in Figure 3b that shows that the folding transition of the MAX1(K15E) variant was intermediately sensitive to pH. Thus, differences in the pH sensitivity as a result of changing the

pH Responsiveness of β-Hairpin Peptide Folding

residue at this position should be easily detected. Arginine (side chain pKa ∼ 12.5) was first incorporated to generate the positive control peptide MAX1(K15R); because this peptide has the same net charge as MAX1 (+9), we expect little difference in its pH-sensitivity compared to MAX1. If large differences are observed, then additional forces other than electrostatics may be playing a role. Figure 4 shows the folding propensity of the peptides listed in Table 2 measured by CD spectroscopy as a function of pH at 50 °C. Here the fraction of β-sheet content is plotted as a function of pH for each of the peptides. The positive control MAX1(K15R) shows little difference compared to MAX1 suggesting that any differences observed for the other peptides should be largely due to electrostatics. Replacing lysine 15 with threonine and isoleucine leads to variants MAX1(K15T) and MAX1(K15I), respectively, each having a net charge of +8 at pH 7. These two peptides behave similarly showing a propensity to fold and assemble at values of pH lower than MAX1; each peptide adopts a moderate fraction of β-sheet structure at pH 8, whereas MAX1 exclusively adopts random coil structure. The fact that these peptides, which contain structurally different residues at position 15, show almost identical pH profiles further supports the assertion that the observed pH-sensitivity is electrostatically driven. Next, the net charge of MAX1 was decreased further to +7 by incorporating Glu resulting in MAX1(K15E); this peptide is capable of folding and assembling at even lower values of pH. Lastly, a peptide containing two glutamic acid residues at positions 4 and 15 was prepared. This peptide has a net charge of +5 and, although we expected it to fold at the lowest solution pH of any of the peptides studied, we were surprised that it is capable of adopting β-sheet structure even at pH 6. These studies indicate that peptides can be designed to fold and assemble at a range of pH values and that there is a direct correlation between the net charge of the peptide and its ability to fold at a given solution pH; for example, peptides with lower net positive charge fold and assemble at lower values of pH. All of the peptides studied thus far contain residue substitutions on the hydrophilic, lysine-rich face of the β-hairpin. The possibility of modulating the net charge of the peptide by making residue substitutions on the hydrophobic face of the hairpin was examined by replacing a valine at position 16 with a glutamic acid resulting in MAX1(V16E). The net charge of this peptide is +8. Figure 5 shows that this peptide is predominantly unfolded at all values of pH studied; for comparison, MAX1(K15E), which has a glutamic acid on the hydrophilic face is shown. The inability of MAX1(V16E) to fold and self-assemble suggests that incorporating a negatively charged residue on the hydrophobic face disrupts the formation of a significant portion of the hydrophobic interactions necessary for these events to take place. Similar phenomena has been observed in sheet rich proteins.54,55 Macroscopic Gelation of MAX1 Variants. The previous CD experiments were performed using 150 µM (∼0.05 wt %) peptide solutions to facilitate the spectroscopic-based measurements of the folding transitions. At this concentration, the peptide folds and self-assembles, but there is not enough assembled peptide to form a mechanically rigid self-supporting hydrogel.48 However, increasing the peptide concentration to 0.5 wt % (∼1.5 mM) results in gel formation when peptide folding and self-assembly is triggered. The kinetics of gelation for the MAX1 variants in Table 2 were investigated both spectroscopically, to monitor the evolution of β-sheet structure, and rheologically, to monitor the evolution of the material’s mechanical rigidity. Although hydrogelation kinetics could be

Biomacromolecules, Vol. 10, No. 9, 2009

2623

measured at multiple values of pH, a common value of pH 8 was chosen for convenience. Although technically demanding, folding and assembly can be followed for gel forming concentrations of peptide by CD. Figure 6a monitors [θ]216 as a function of time for 0.5 wt % solutions of peptide at pH 8. MAX1 and MAX1(K15R) were incapable of folding at this pH at lower peptide concentrations, however, at 0.5 wt % they slowly fold and assemble. This is not surprising because, mechanistically, peptide folding and selfassembly are most likely linked equilibria and self-assembly is concentration dependent. Comparing the data for all the peptides in Figure 6a demonstrates that the rate at which these peptides fold and assemble into hydrogel is dependent on their net charge. Peptides having lower net positive charge, fold and assemble more quickly than peptides of higher net positve charge. Hydrogelation kinetics were also assessed rheologically. Figure 6b shows the evolution of storage moduli (G′), a measure of the gel’s mechanical rigidity, with respect to time. Here gelation, as a result of peptide folding and assembly, of 0.5 wt % peptide solutions is initiated by ramping the temperature from 5 to 50 °C. The rheological data mirror the spectroscopic data in that peptides having lower net positive charge form gels more quickly than gels of higher positive charge. For example, MAX1(K15E) (net charge ) +7) undergoes gelation much more quickly than MAX1 (net charge ) +9). MAX1(K15I) and MAX1(K15T) have intermediate net charges of +8, and correspondingly, form gels at rates intermediate to MAX1 and MAX1(K15E). The control peptide, MAX1(K15R), having the same net charge as MAX1 undergoes gelation very slowly. The +5 charged peptide MAX1(K4E, K15E), which one would predict should form hydrogel the fastest, only precipitated from solution, making gelation studies impossible. Inspection of Figure 6b also shows that these peptides form gels of differing mechanical rigidity. In general, differences in the rigidity of gels composed of semiflexible fibrils can be attributed to the number of cross-links present within the system. It has been shown that, for gels formed with semiflexible polymers such as actin, small changes in the cross-link density at constant volume fractions can dramatically alter the elastic modulus over a few orders of magnitude.56,57 In contrast to actin gels, where cross-links are formed by actin binding proteins, cross-links are formed in the hairpin-based gels as a consequence of the self-assembly event. The rheological data in Figure 6b indicate that peptides that fold and assemble more quickly, ultimately form more mechanically rigid gels. For example, MAX1(K15E) forms a gel having a storage modulus of nearly 600 Pa compared to the 50 Pa afforded by the MAX1 gel. Gels formed from the +8 charged peptides are of intermediate rigidity. The physical nature of the cross-links formed in these gels and how assembly kinetics influences cross-link density are areas of active investigation and will be reported in due course. Local Fibril Morphology of MAX1 Variants. Extensive transmission electron microscopy (TEM) and small angle neutron scattering (SANS) studies indicate that MAX1 selfassembles into a highly cross-linked network of β-sheet rich fibrils.45,46 Fibrils are composed of a bilayer of hairpins that intermolecularly hydrogen bond to form the long axis of an individual fibril, Figure 1. The width of a given fibril is defined by the width of an individual hairpin, namely, 3 nm. Unlike many reported peptide-based fibrils,35,38,58 MAX1 fibrils do not show a propensity to form higher order assemblies such as laminates. TEM is used here to study the local morphologies of the fibrils formed by the MAX1 variants. Figure 7 shows

2624

Biomacromolecules, Vol. 10, No. 9, 2009

Rajagopal et al.

Figure 7. Transmission electron microscopy image showing the nanoscale structure of the fibrils for MAX1 (a) and MAX1(K15E) (b). Width of the fibrils indicated by the arrows is homogeneously uniform (approximately 3 nm).

Figure 8. Folding, self-assembly, and hydrogelation kinetics for 0.5 wt % MAX 1 and MAX1(K15E) at pH 7.4, 150 mM NaCl, and 37 °C. (a) β-Sheet formation is monitored for both peptides by measuring [θ]216 as a function of time. (b) The rates are measured by monitoring the evolution of storage moduli with time.

TEM images for MAX1 and MAX1(K15E) fibrils. It is evident that the local morphology of the variant is nearly identical to MAX1; fibrils are monodisperse in width (∼3 nm) and there is little sign of lamination. Fibrils formed from the other peptide variants in Table 2 display similar morphologies (see Supporting Information). Taken together, the TEM experiments suggest that although the rate of folding and self-assembly leading to hydrogelation is different for each peptide variant, they form similar supramolecular structures. This data also suggests that making residue substitutions on the hydrophilic face of the hairpin does not influence the local morphology of the fibrils that constitute the gel. Gelation at Physiological Conditions. By understanding how hairpin electrostatics influences hydrogelation, it is possible to formulate peptides for specific applications. For example, hydrogels that can be formed near physiological conditions (pH 7.4, 150 mM NaCl, and 37 °C) may find use in biological applications such as in vivo material formation and cellular encapsulation.47 The data in Figure 4 shows that MAX1(K15E) is capable of folding and assembling to an intermediate degree at pH 7.0 when micromolar quantities are dissolved in a solution of low ionic strength (10 mM NaCl). We hypothesized that if millimolar quantities (0.5 wt %) of the peptide were dissolved in solutions of higher ionic strength (150 mM NaCl), then MAX1(K15E) could undergo efficient hydrogelation at 37 °C. Increasing the ionic strength should help screen the electrostatic

repulsion experienced as the peptide folds, and increasing the peptide concentration should hasten the rate of self-assembly, affording a mechanically rigid gel. Figure 8a shows the rate of β-sheet formation of MAX1(K15E) versus MAX1 at 0.5 wt % at pH 7.4, 150 mM NaCl, and 37 °C. The data show that the variant folds and assembles within the first few seconds of the experiment, whereas MAX1 completes its transitions only after about 60 min. Importantly, Figure 8b shows the evolution of the storage moduli for the two gels. MAX1(K15E) forms a more mechanically rigid gel more quickly than MAX1. In fact, at 0.5 wt % under these conditions, MAX1 forms only a very weak gel, whereas the variant forms a moderately rigid gel after about a minute.

Conclusions Structure-based design was used to generate a suite of peptides capable of undergoing thermally induced gelation at distinct values of pH. Point substitutions of the lysine residues of the self-assembling peptide MAX1 were made to alter the net charge of the peptide. Peptides of lower net positive charge are capable of folding and assembling into hydrogel material at lower values of pH at a given temperature. The pH-sensitive folding and assembling behavior is not only dependent on the net peptide charge but also on the exact position of substitution within the peptide sequence. Mechanically, the bulk material

pH Responsiveness of β-Hairpin Peptide Folding

rigidity of the gels prepared from these peptides is dependent on the rate of hydrogel formation. Peptides that fold and assemble faster afford more rigid gels. This design strategy yielded a peptide MAX1(K15E) that is capable of undergoing thermally triggered hydrogelation at physiological buffer conditions (pH 7.4, 150 NaCl, 37 °C). Acknowledgment. This work was supported by NSF CHE0348323 (J.P.S.) and NSF Career DMR 0348147. Supporting Information Available. Representative analytical HPLC chromatograms, ESI-MS spectra of purified peptides, dilute and concentrated solution CD spectra, oscillatory rheology data of hydrogels, and TEM images of the peptides in Table 2. This material is available free of charge via the Internet at http:// pubs.acs.org.

References and Notes (1) Lee, K. Y.; Mooney, D. J. Chem. ReV. 2001, 101, 1869–1879. (2) Kikuchi, A.; Okano, T. AdV. Drug DeliVery ReV. 2002, 54, 53–77. (3) Eddington, D. T.; Beebe, D. J. AdV. Drug DeliVery ReV. 2004, 56, 199–210. (4) Peppas, N. A.; Hilt, J. Z.; Khademhosseini, A.; Langer, R. AdV. Mater. 2006, 18, 1345–1360. (5) Peppas, N. A.; Bures, P.; Leobandung, W.; Ichikawa, H. Eur. J. Pharm. Biopharm. 2000, 50, 27–46. (6) Hoffman, A. S. AdV. Drug DeliVery ReV. 2002, 54, 3–12. (7) Drury, J. L.; Mooney, D. J. Biomaterials 2003, 24, 4337–4351. (8) Byrne, M. E.; Park, K.; Peppas, N. A. AdV. Drug DeliVery ReV. 2002, 54, 149–161. (9) Alarcon, C. d. l. H.; Pennadam, S.; Alexander, C. Chem. Soc. ReV. 2005, 34, 276–285. (10) Sorber, J.; Steiner, G.; Schulz, V.; Guenther, M.; Gerlach, G.; Salzer, R.; Arndt, K.-F. Anal. Chem. 2008, 80, 2957–2962. (11) Richter, A.; Paschew, G.; Klatt, S.; Lienig, J.; Arndt, K.-F.; Adler, H.-J. P. Sensors 2008, 8, 561–581. (12) Dhanarajan, A. P.; Siegel, R. A. Macromol. Symp. 2005, 227, 105– 114. (13) Yu, Y.-q.; Xu, Y.; Ning, H.; Zhang, S.-s. Colloid Polym. Sci. 2008, 286, 1165–1171. (14) Musch, J.; Schneider, S.; Lindner, P.; Richtering, W. J. Phys. Chem. B 2008, 112, 6309–6314. (15) Klouda, L.; Mikos, A. G. Eur. J. Pharm. Biopharm. 2008, 68, 34–45. (16) Fang, J.-Y.; Chen, J.-P.; Leu, Y.-L.; Hu, J.-W. Eur. J. Pharm. Biopharm. 2008, 68, 626–636. (17) Matsumoto, S.; Yamaguchi, S.; Ueno, S.; Komatsu, H.; Ikeda, M.; Ishizuka, K.; Iko, Y.; Tabata, K. V.; Aoki, H.; Ito, S.; Noji, H.; Hamachi, I. Chem.sEur. J. 2008, 14, 3977–3986. (18) Luo, Y.; Shoichet, M. S. Nat. Mater. 2004, 3, 249–253. (19) Gattas-Asfura, K. M.; Weisman, E.; Andreopoulos, F. M.; Micic, M.; Muller, B.; Sirpal, S.; Pham, S. M.; Leblanc, R. M. Biomacromolecules 2005, 6, 1503–1509. (20) Fairman, R.; Aakerfeldt, K. S. Curr. Opin. Struct. Biol. 2005, 15, 453– 463. (21) Hamley, I. W. Angew. Chem., Int. Ed. 2007, 46, 8128–8147. (22) Holmes, T. C. Trends Biotechnol. 2002, 20, 16–21. (23) Mart, R. J.; Osborne, R. D.; Stevens, M. M.; Ulijn, R. V. Soft Matter 2006, 2, 822–823. (24) Rajagopal, K.; Schneider, J. P. Curr. Opin. Struct. Biol. 2004, 14, 480–486. (25) Zhang, S. Nat. Biotechnol. 2003, 21, 1171–1178. (26) Ellis-Behnke, R. G.; Liang, Y.-X.; You, S.-W.; Tay, D. K. C.; Zhang, S.; So, K.-F.; Schneider, G. E. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 5054–5059. (27) Kisiday, J.; Jin, M.; Kurz, B.; Hung, H.; Semino, C.; Zhang, S.; Grodzinsky, A. J. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 9996–10001.

Biomacromolecules, Vol. 10, No. 9, 2009

2625

(28) Davis, M. E.; Hsieh, P. C. H.; Takahashi, T.; Song, Q.; Zhang, S.; Kamm, R. D.; Grodzinsky, A. J.; Anversa, P.; Lee, R. T. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 8155–8160. (29) Nagai, Y.; Unsworth, L. D.; Koutsopoulos, S.; Zhang, S. J. Controlled Release 2006, 115, 18–25. (30) Papapostolou, D.; Smith, A. M.; Atkins, E. D. T.; Oliver, S. J.; Ryadnov, M. G.; Serpell, L. C.; Woolfson, D. N. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 10853–10858. (31) Woolfson, D. N.; Ryadnov, M. G. Curr. Opin. Chem. Biol. 2006, 10, 559–567. (32) Smith, A. M.; Banwell, E. F.; Edwards, W. R.; Pandya, M. J.; Woolfson, D. N. AdV. Funct. Mater. 2006, 16, 1022–1030. (33) MacPhee, C. E.; Woolfson, D. N. Curr. Opin. Solid State Mater. Sci. 2004, 8, 141–149. (34) Rughani, R. V.; Schneider, J. P. MRS Bull. 2008, 33, 530–535. (35) Lu, K.; Jacob, J.; Thiyagarajan, P.; Conticello, V. P.; Lynn, D. G. J. Am. Chem. Soc. 2003, 125, 6391–6393. (36) Isenberg, H.; Kjaer, K.; Rapaport, H. J. Am. Chem. Soc. 2006, 128, 12468–12472. (37) De la Paz, M. L.; Goldie, K.; Zurdo, J.; Lacroix, E.; Dobson, C. M.; Hoenger, A.; Serrano, L. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 16052–16057. (38) Aggeli, A.; Nyrkova, I. A.; Bell, M.; Harding, R.; Carrick, L.; McLeish, T. C. B.; Semenov, A. N.; Boden, N. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 11857–11862. (39) Zhao, Y.; Yokoi, H.; Tanaka, M.; Kinoshita, T.; Tan, T. Biomacromolecules 2008, 9, 1511–1518. (40) Goeden-Wood, N. L.; Keasling, J. D.; Muller, S. J. Macromolecules 2003, 36, 2932–2938. (41) Collier, J. H.; Hu, B.-H.; Ruberti, J. W.; Zhang, J.; Shum, P.; Thompson, D. H.; Messersmith, P. B. J. Am. Chem. Soc. 2001, 123, 9463–9464. (42) Caplan, M. R.; Schwartzfarb, E. M.; Zhang, S.; Kamm, R. D.; Lauffenburger, D. A. Biomaterials 2001, 23, 219–227. (43) Aggell, A.; Bell, M.; Bdoen, N.; Keen, J. N.; Knowles, P. F.; McLeish, T. C. B.; Pitkeathly, M.; Radford, S. E. Nature 1997, 386, 259–262. (44) Ozbas, B.; Kretsinger, J.; Rajagopal, K.; Schneider, J. P.; Pochan, D. J. Macromolecules 2004, 37, 7331–7337. (45) Schneider, J. P.; Pochan, D. J.; Ozbas, B.; Rajagopal, K.; Pakstis, L.; Kretsinger, J. J. Am. Chem. Soc. 2002, 124, 15030–15037. (46) Ozbas, B.; Rajagopal, K.; Schneider, J. P.; Pochan, D. J. Phys. ReV. Lett. 2004, 93, 268106/1–268106/4. (47) Haines-Butterick, L.; Rajagopal, K.; Branco, M.; Salick, D.; Rughani, R.; Pilarz, M.; Lamm, M. S.; Pochan, D. J.; Schneider, J. P. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 7791–7796. (48) Veerman, C.; Rajagopal, K.; Palla, C. S.; Pochan, D. J.; Schneider, J. P.; Furst, E. M. Macromolecules 2006, 39, 6608–6614. (49) Privalov, P. L.; Makhatadze, G. I. J. Mol. Biol. 1993, 232, 66079. (50) Privalov, P. L. Crit. ReV. Biochem. Mol. Biol. 1990, 25, 281–305. (51) Murphy, K. P.; Privalov, P. L.; Gill, S. J. Science 1990, 247, 55961. (52) Baldwin, R. L. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 8069–72. (53) The pKa of lysine side chain is approximately 10.5. The modified electrostatic environment due to the presence of the 8 lysines within the MAX1 sequence is expected to perturb the pKa such that a small proportion are deprotonated even at pH 9, including the N-terminal amine (pKa ) 9.2). (54) Richardson, J. S.; Richardson, D. C. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2754–2759. (55) Wang, W. X.; Hecht, M. H. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2760–2765. (56) Levine, A. J.; Head, D. A.; MacKintosh, F. C. J. Phys.: Condens. Matter 2004, 16, S2079-S2088. (57) Gardel, M. L.; Shin, J. H.; MacKintosh, F. C.; Mahadevan, L.; Matsudaira, P.; Weitz, D. A. Science 2004, 304, 1301–1305. (58) Lamm, M. S.; Rajagopal, K.; Schneider, J. P.; Pochan, D. J. J. Am. Chem. Soc. 2005, 127, 16692–16700.

BM900544E