Contribution of Electrostatics in the Fibril Stability of a Model Ionic

Nov 23, 2015 - The results of this work indicate that the global peptide net charge is a key property that correlates well with the fibril stability, ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/Biomac

Contribution of Electrostatics in the Fibril Stability of a Model IonicComplementary Peptide Marta Owczarz, Tommaso Casalini, Anna C. Motta, Massimo Morbidelli,* and Paolo Arosio† Institute for Chemical and Bioengineering, Department of Chemistry and Applied Biosciences, ETH Zurich, 8093 Zurich, Switzerland S Supporting Information *

ABSTRACT: In this work we quantified the role of electrostatic interactions in the self-assembly of a model amphiphilic peptide (RADA 16-I) into fibrillar structures by a combination of size exclusion chromatography and molecular simulations. For the peptide under investigation, it is found that a net charge of +0.75 represents the ideal condition to promote the formation of regular amyloid fibrils. Lower net charges favor the formation of amorphous precipitates, while larger net charges destabilize the fibrillar aggregates and promote a reversible dissociation of monomers from the ends of the fibrils. By quantifying the dependence of the equilibrium constant of this reversible reaction on the pH value and the peptide net charge, we show that electrostatic interactions contribute largely to the free energy of fibril formation. The addition of both salt and a charged destabilizer (guanidinium hydrochloride) at moderate concentration (0.3−1 M) shifts the monomer-fibril equilibrium toward the fibrillar state. Whereas the first effect can be explained by charge screening of electrostatic repulsion only, the promotion of fibril formation in the presence of guanidinium hydrochloride is also attributed to modifications of the peptide conformation. The results of this work indicate that the global peptide net charge is a key property that correlates well with the fibril stability, although the peptide conformation and the surface charge distribution also contribute to the aggregation propensity.



applications in 3D cell cultures,12 tissue regeneration and engineering,13 as well as in drug release14 and drug delivery.15,16 Both the kinetic and the thermodynamic stability of the fibrillar structures are regulated by a series of intermolecular forces, which can be tuned by modifying several operating parameters. Amyloid fibrils are characterized by a high content of cross-βsheet structure organized into thin, long, and unbranched fibrils.17,18 The network of hydrogen bonds in the peptide backbone plays a crucial role in stabilizing the amyloid structure.19 Indeed, under suitable conditions, a large number of peptides and proteins are able to form amyloid fibrils that share similar structure independently of the protein primary sequence.2,20 In addition to hydrogen bonding,19 several other intermolecular forces contribute to the fibril formation and stability, including aromatic21,22 and electrostatic interactions. Steric zippers formed by facing side chains of pair β-sheets23 and curvature energies related to the twisting of the fibrils also play an important role, in particular in multistranded filaments.24−27 Within this complex scenario, understanding the role of electrostatics in the self-assembly of peptides and proteins is a difficult task. This challenge arises largely from the heterogeneous distribution of charges on the protein surface, which

INTRODUCTION

The self-assembly of peptides and proteins into regular fibrillar structures known as amyloid fibrils is involved in a large variety of research fields ranging from biological to material sciences. For instance, amyloid fibrils are increasingly recognized to be involved in the onset and the progression of several neurodegenerative diseases such as Alzheimer’s, Parkinson’s, and Huntington’s disease.1,2 Not only dysfunctional but also functional amyloids are observed in nature. An intriguing example is the storage of certain peptides, for instance hormones, as amyloid fibrils, which release the active monomeric form of the peptides in a controlled dissociation process.3 This naturally occurring form of storing active biomolecules guarantees high mechanical and chemical stability, as well as the ability to release the active component in a controlled way, and may inspire formulations of long-acting drugs.4 In the context of biotechnology, several small peptides have been rationally developed to spontaneously self-assemble in aqueous solutions into higher ordered structures depending on the environmental conditions.5−10 One particular class of these functional materials is the family of ionic-complementary peptides, which exhibit hydrophobic residues on one side of the structure and charged amino acids on the other side.6,11 The controlled aggregation into chemically and mechanically resistant structures makes these peptides ideal building-blocks for the formation of biocompatible materials, which found © 2015 American Chemical Society

Received: August 13, 2015 Revised: November 12, 2015 Published: November 23, 2015 3792

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules

Transmission Electron Microscopy (TEM). Peptide samples were imaged using a FEI Morgagni 268 microscope. Peptide solutions were diluted to a final concentration of about 0.05 g/L, and loaded on a carbon grid (Quantifoil, Jena, Germany). Samples at pH 2.0 and 4.5 were stained with a 2% uranyl acetate solution, whereas samples at pH ranging from 6.0 to 8.5 were stained with a 2% phosphotungstic acid solution (pH 6.5). Size Exclusion Chromatography (SEC). Size exclusion chromatography analysis was performed using a Superdex Peptide 10/300 GL, 10 mm × 300 mm size-exclusion column (GE Healthcare, Uppsala, Sweden) mounted on a Agilent 1100 series HPLC unit (Santa Clara, CA, USA) consisting of an isocratic pump with degasser, an autosampler, a column oven, and a DAD detector. Each sample was eluted for 70 min at a constant flow rate of 0.4 mL/min using a buffer at suitable pH as mobile phase. The UV absorbance peaks were detected at 217 nm.

involves the simultaneous presence of both positive and negative charges that are highly dependent on the environmental pH and buffer composition. 28−36 Despite this heterogeneous distribution of charges on the protein surface, the global surface net charge has been recognized to be a crucial physicochemical parameter that determines the general aggregation behavior.37 For instance, close to the isoelectric point (pI), corresponding to zero net charge, peptides and proteins tend to precipitate into amorphous aggregates, whereas higher net charges promote the formation of fibrillar structures.38,39 In this study, we provide a more quantitative framework of this qualitative behavior by evaluating the contribution of electrostatics in the free energy of fibril formation, considering the well-known amphiphilic peptide RADARADARADARADA (RADA 16-I) as a model system. We apply size exclusion chromatography, circular dichroism, and electron microscopy to characterize the aggregate structure and quantify the monomer content at different pH values. In order to gain insight into the structural conformation of the monomeric peptide at different pH values, we complement the experimental characterization with metadynamics simulations, which have been proved to provide an efficient sampling of protein structures,40,41 even in the presence of denaturant.42,43 We show that the global peptide net charge is a key property that correlates well with the fibril stability, although the peptide conformation and the surface charge distribution also contribute to the self-assembly propensity.





COMPUTATIONAL METHODS Guanidinium Atomic Charges. The guanidinium structure was optimized in vacuo by means of density functional theory (DFT) calculations at B3LYP/6-31G(d,p) level of theory.45−47 The obtained geometry was further optimized in implicit water and modeled through the integral equation formalism polarizable continuum model (IEFPCM)48 at 300 K, at the same level of theory. Atomic charges were fitted by means of RESP formalism,49,50 starting from the electrostatic potentials computed through DFT calculations at B3LYP/6-311+G(d,p)51 level of theory in implicit water. Charge fitting was performed in two steps: first, atomic charges were determined assigning an overall charge equal to +1. In a second step, charge equivalence for chemically equivalent atoms was imposed. In particular, all hydrogen and nitrogen atoms have the same partial atomic charge because of resonance structures. Computations were performed by means of Gaussian 09 software.52 Simulations Protocol. Metadynamics simulations were performed following the approach proposed by Deighan et al.,53 who combined metadynamics and parallel tempering (PTMetaD) in the well-tempered ensemble (PTMetaD-WTE) in order to improve simulation efficiency. Metadynamics simulations allow the recovery of the free energy of a system by keeping track and discouraging the sampling of the conformations already visited during the simulations as a function of few conformational parameters (collective variables).54,55 All simulations were carried out using GROMACS 5.0.256 patched with PLUMED 2.1.0 plugin.57 Linear RADA 16-I peptide structures were built using tleap module as implemented in AmberTools package;58 protonation states at pH 2.0 and pH 4.5 were determined according to PropKa calculations.59 At pH 2.0, all arginine and aspartic acid side chains are protonated, thus leading to an overall net charge equal to +4; at pH 4.5, all arginine side chains are protonated, and three single aspartic acid moieties are dissociated, and the net charge is equal to +1. According to PropKa calculations, in the protein primary structure all aspartic acid residues exhibit the same pKa value. Although it can be expected that in the protein arrangement a dynamic exchange of proton among carboxyl groups takes place, a reliable constant pH method (needed to properly account for this dynamic equilibrium) is challenging to implement in GROMACS. Therefore, in the present simulations at pH 4.5, the protonation has been arbitrarily assigned to the first aspartic acid moiety in peptide sequence. Simulations were performed adopting ff14SB force

EXPERIMENTAL SECTION

Material and Sample Preparation. RADA 16-I peptide (Ac-R-AD-A-R-A-D-A-R-A-D-A-R-A-D-A-CONH2 with MW 1712.8 Da and purity ∼80% as assessed by mass spectrometry analysis as shown in Figure S1 in the Supporting Information) was provided by Lonza Ltd. (Visp, Switzerland) as lyophilized powder in the form of trifluoroacetic salt. The material was used without further purification. Before each experiment, a stock solution of the peptide at a concentration of 5 g/L was freshly prepared by dissolving the peptide powder in 10 mM HCl at pH 2.0 following the protocol described elsewhere.44 Briefly, a suitable amount of buffer was added to the weighted peptide powder, and the solution was gently stirred at room temperature for 10 min for homogenization. The formation of fibrils was observed immediately after solubilization of the powder. Samples for analysis were prepared either by diluting 10-fold the stock solution in suitable buffers, or by dialyzing the stock solution against suitable buffers using the DispoBiodialyzer kit with 1 kDa molecular weight cutoff membrane (SigmaAldrich GmbH, Steinheim, Germany). Analysis of selected samples showed no significant difference between solutions prepared according to the two different applied protocols (data not shown). In order to guarantee proper buffering, the following solutions were used: 10 mM acetate buffer at pH from 3.0 to 5.0, 10 mM phosphate buffer at pH 6.0 and 7.0, and 10 mM TRIS-HCl at pH 8.5. The final pH value of the samples was checked by pH-Fix 0−14 color-fixed indicator sticks (Machery-Nagel GmbH & Co. KG, Düren, Germany). Circular Dichroism (CD). Circular dichroism measurements were performed with a Jasco-815 CD spectrophotometer (Jasco, Easton, MD, USA). CD spectra were recorded in the far-UV region from 190 to 260 nm at 25 °C. A quartz cuvette with 0.1 cm path length was used. Spectra obtained after buffer subtraction were corrected and smoothed using the Savitsky-Golay function. Freshly prepared samples of RADA 16-I at 5 g/L in 10 mM HCl at pH 2.0 were diluted in 10 mM HCl (pH 2.0), 10 mM acetate buffer (pH 4.5), 10 mM phosphate buffer (both pH 6.0 and 7.0), and 10 mM TRIS-HCl buffer (pH 8.5) in order to obtain a peptide final concentration of 0.5 g/L. The pH value after sample dilution was checked by pH-Fix 0−14 color-fixed indicator sticks (Machery-Nagel GmbH & Co. KG, Düren, Germany). 3793

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules

Figure 1. pH effect on the secondary structure and the morphology of RADA 16-I peptide. (a) CD spectra of 0.5 g/L peptide solutions in 10 mM HCl at pH 2.0 (blue dashed line), 10 mM acetate buffer at pH 4.5 (red continuous line), 10 mM phosphate buffer at pH 6.0 (green dotted line), and 7.0 (violet dash-dotted line) and 10 mM TRIS-HCl buffer at pH 8.5 (black dashed line). (b−f) TEM pictures of the RADA 16-I solutions in 10 mM HCl at pH 2.0 (b), 10 mM acetate buffer at pH 4.5 (c), 10 mM phosphate buffer at pH 6.0 (d), 10 mM phosphate buffer at pH 7.0 (e), and 10 mM TRIS-HCl buffer at pH 8.5 (f).

field60 for RADA 16-I peptide and General Amber Force Field (GAFF)61 for guanidinium. Each peptide was solvated using 5500 TIP3P water molecules62 for simulations in water environment, or 7000 TIP3P water molecules and 80 guanidinium molecules when 0.6 M guanidinium hydrochloride solution was simulated. Electroneutrality was assured by adding Cl− ions, whose parameters (optimized for TIP3P water model) has been taken from Joung and Cheatham.63 The starting system for PTMetaD-WTE simulations was obtained using the following protocol. First, energy minimization was performed by means of conjugate gradient method in order to remove bad solvent/ solvent and solvent/solute contacts. Temperature was raised from 0 to 300 K through 200 ps in NVT ensemble, applying a weak harmonic restraint on solute molecules in order to avoid wild fluctuation. The system was finally equilibrated with 1 ns simulation in NPT ensemble at 300 K and 1 atm. Velocity rescale algorithm64 and Parrinello−Rahman barostat65 were employed to maintain the system at the desired values of temperature and pressure, respectively. Electrostatic long-range interaction were treated through the particle mesh Ewald (PME) method66 using a cutoff value equal to 12 Å; the same cutoff was used for Lennard-Jones interactions. The neighbor list was updated every 5 fs, and all covalent bonds involving hydrogen were restrained by means of the LINCS algorithm.67 PTMetaD-WTE simulations were carried out using 12 replicas whose temperatures range from 300 to 631 K. Before starting metadynamics simulations, a 1 ns molecular dynamics simulation was performed for each replica. In this way, every system could equilibrate according to its temperature, and adjacent replicas exhibit different initial peptide structures. PTMetaD-WTE simulations were then performed using a two-step protocol.53 First, a 5 ns PTMetaD simulation41 was carried out biasing only potential energy, chosen as a collective variable; Gaussians were deposited every 0.5 ps, adopting a height value equal to 0.28 kcal/mol, a width value equal to 500 kcal/mol, and a bias factor equal to 24. In a second step, the radius of gyration related to α-carbon atoms in the protein backbone (Rg) and the intrachain hydrogen bonds (Hb) were

chosen as collective variables (since they proved to be suitable for sampling protein secondary structures40,41,43) and biased according to well-tempered metadynamics algorithm.55 No more Gaussians were added in order to bias potential energy; the cumulative bias was used as a static additional bias potential. Focusing on the radius of gyration and the intrachain hydrogen bonds, Gaussians were added every 1 ps, adopting a height value equal to 0.28 kcal/mol, width values equal to 0.01 and 0.1 for Rg and Hb respectively and a bias factor equal to 8. The velocity rescale algorithm64 was used in order to maintain the systems at the desired temperature values. Long−range interactions were treated as described above. A simulation time equal to 250 ns was used for each replica, thus leading to a total simulation time of 3 μs for each system. Exchange attempts between replicas were performed every 0.4 ps by means of the PTMetaD algorithm as implemented in GROMACS 5.0.2 patched with PLUMED 2.1.0 plugin. Convergence was verified by evaluating the folding free energy as a function of simulated time; details are reported in Figure S2 in the Supporting Information. Solvent accessible surface area (SASA) was calculated through the g_sas utility as implemented in the GROMACS package. Electrostatic potentials have been computed through Adaptive Poisson−Boltzmann Solver (APBS).68 Electrostatic potential is expressed in kBTe−1 units, where kB is the Boltzmann constant, T is the absolute temperature, and e is the electron charge.



RESULTS AND DISCUSSION pH Effect on the RADA 16-I Secondary Structure and Morphology. We start our analysis by investigating the effect of pH on the secondary structure of the peptide to compare our results with the broad range of data on amphiphilic peptides reported in several works in the literature.6,37,69,70 The effect of the environmental pH on the peptide secondary structure was determined by circular dichroism analysis (Figure 1a). At pH 2.0, the minimum at 195 nm in the CD spectrum clearly indicates the presence of a large amount of random coil structures, which can be attributed to the monomeric peptide, 3794

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules whereas the shoulder visible at the wavelength of about 215 nm corresponds to the β-sheet structures of the fibrils. At pH 4.5, we observe mainly β-sheet structures, representative of a large amount of fibrils. An increase of the pH of the solution to neutral conditions (pH 6.0, 7.0 and 8.5) leads to the loss of βsheet structures. Moreover, the formation of amorphous precipitates in the micron size range was detected by macroscopic observations. In parallel with CD analysis, we investigated the morphology of the RADA 16-I peptide by transmission electron microscopy. We observed the presence of long, thin and rather rigid fibrils at pH 2.0 and 4.5 (Figure 1b,c), and a mixture of fibrils and amorphous precipitates at higher pH (Figure 1d−f). Our results are in excellent agreement with the findings reported for RADA 16-I by Ye et al. and Zhang et al.69,70 Effect of Electrostatic Interactions on the MonomerFibril Equilibrium. The analysis of the secondary structure content of RADA 16-I peptide at different pH values indicates that the monomer fraction in the dispersions of RADA 16-I peptide at 0.5 g/L is predominant at pH 2.0, whereas at pH 4.5 the CD spectrum shows mainly the presence of fibrils (Figure 1a). We quantified in detail the dependence of the monomer amount on the pH value by measuring the monomeric peptide concentration in freshly prepared solutions of RADA 16-I at 0.5 g/L in the pH range 2.0−5.0 by size exclusion chromatography. We observed a decrease in the monomer percentage with increasing the pH from pH 2.0 to pH 4.5. Since precipitation of amorphous aggregates is observed at pH 5.0, as shown in Figure 2a, this analysis indicates that the optimal pH for the self-assembly of monomers into fibrils is equal to about 4.5. The monomer content at different pH values can be correlated to the net charge of the peptide calculated from the amino acid composition and the pKa values of the single amino acids, as described elsewhere.71 At low pH value, where the largest amount of monomer is observed, the high amount of uncompensated charge induces repulsive interactions between monomer molecules, thus disfavoring the aggregation process and destabilizing the fibrils. The decrease in the net charge by increasing the pH value results in the increase of the fibril fraction in the solution, whereas close to the isoelectric point, which is equal to about 7.2,69 the absence of stabilizing electrostatic interactions induces the precipitation of amorphous aggregates in the micron size range (Figure 2a). Actually, precipitates are already formed at pH 5.0 (star symbol in Figure 2a). Although the net charge is not equal to zero, it is still not sufficient for stabilization. The maximum pH, at which the precipitation was not observed and the maximum aggregation propensity of monomers into fibrils occurred, is equal to 4.5, corresponding to a net charge equal to +0.75. This result is in excellent agreement with the value of ±1 reported by López de la Paz et al.72 as well as Aggeli et al.37 as the optimum net charge for amyloid fibril formation, indicating the possible presence of a certain level of universality in the aggregation behavior of peptides. The effect of the surface net charge on the aggregation propensity of the peptide was further rationalized by performing metadynamics simulations of the structures of the monomeric peptides in solution. In particular, the monomer structure was calculated at two selected reference pH values, namely, pH 2.0 and 4.5, at which, respectively, the minimum and the maximum of the monomer aggregation extent into fibrils were observed (Figure 2b−g). The number of intrapeptide hydrogen bonds and the radius of gyration of α-carbon

Figure 2. Effect of pH and net charge on the peptide stability and the aggregate morphology. (a) Fractions of monomeric peptide (⧫) and fibrils (●) as a function of the solution pH and the peptide net charge, zp, for RADA 16-I solutions at peptide concentration of 0.5 g/L. The red star corresponds to the critical pH = 5.0 and net charge zp = 0.27 at which precipitation of amorphous aggregates is observed. If error bars are not visible, they are smaller than the symbol. (b,c) Free energy surface calculated from PTMetaD-WTE simulations at pH 4.5 (b) and pH 2.0 (c); contour lines are plotted every kBT unit (2.5 kJ/mol). (d,e) Reference minimum energy structures at pH 4.5 (d) and pH 2.0 (e). (f,g) Electrostatic surface potential expressed in kBTe−1 units at pH 4.5 (f) and pH 2.0 (g). Representative SEC chromatograms are shown in Figure S3 in the Supporting Information.

atoms in the peptide backbone were used as collective coordinates to determine the minimum free energy surface under the considered conditions. In particular, in Figure 2b,c it is seen that such minimum occurs at a radius of gyration of about 0.56 nm at pH 2.0 and about 0.66 nm at pH 4.5. According to the simulations, the global minimum of the free energy surface at pH 4.5 (which implies a low net charge value) corresponds to β-hairpin-like structures of the monomer (Figure 2d). Both positive and negative charges are exposed on monomer surface, thus promoting an electrostatic-driven aggregation between peptides (Figure 2f). More ordered βhairpin structures as well as unordered open structures can be found in the neighbor regions of the minimum (ΔF ≤ 6 kJ/ mol). At pH 2.0, the global minimum of the free energy surface corresponds to compact random coil structures of the monomeric peptide, whose mean radius of gyration is equal to 0.56 nm. More open random coil structures, corresponding to a radius of gyration up to 0.67 nm are contained in the neighborhood of the minimum (ΔF ≤ 4 kJ/mol). In all sampled structures, the positively charged arginine side chains 3795

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules are located on the peptide surface and are fully exposed toward the water environment (Figure 2e), in order to maximize the favorable interactions between water molecules and the charged moieties. The resulting average surface potential is thus highly positive (blue-colored areas in Figure 2g), confirming the presence of a repulsive electrostatic barrier between two interacting monomers, which hinders the fibril formation. The structures obtained by metadynamics simulations (Figure 2d−f) are in good agreement with both the secondary structure content measured by CD spectroscopy (Figure 1a) and the experimental observation of the aggregation behavior. Indeed, the β-hairpin-like structure at pH 4.5 along with the low net charge is likely to promote fibril formation. By contrast, at pH 2.0 the aggregation is hindered by both electrostatic repulsion and the random coil geometry of the monomeric units. After quantifying the monomer amount as a function of the pH value, we verified the presence of equilibrium between monomers and fibrils by measuring the dependence of the amount of free monomer as a function of the total peptide concentration. The fibril formation can be described as a linear polymerization process consisting of a series of reversible additions of monomers (M) to fibrils with length n − 1 (Fn−1) to generate fibrils with length n (Fn):73 Fn − 1 + M ↔ Fn

(1) Figure 3. Monomer−fibril equilibrium. (a) The effect of the total peptide concentration on the monomer content of RADA 16-I in 10 mM HCl at pH 2.0 (⧫, dashed line) and pH 2.5 (■, dotted line), 10 mM acetate buffer at pH 3.0 (▲, dash-dotted line) and 4.0 (●, solid line). The lines represent the basic linear fitting of the experimental points according to eq 6. The inset is the zoom-in of the results at pH 2.0 and pH 2.5. (b) pH and net charge, zp, effect on the monomer− fibril equilibrium constant of RADA 16-I peptide. If error bars are not visible, they are smaller than the symbol. Representative SEC chromatograms are shown in Figure S4 in the Supporting Information.

The equilibrium constant of this reversible association reaction of monomers to the fibril ends can be defined as Keq =

[Fn] [Fn − 1]·[M ]

(2)

where [Fn], [Fn−1], and [M] are the concentrations of fibrils with length n, fibrils with length n − 1, and monomer, respectively. The model shown here is not accounting for the short fibrils (i.e., small n), but the approximation is valid for the long fibrils present in our system. As described elsewhere,73 the total peptide concentration (Ctot) is equal to ∞

C tot =



∑ n·[Fn] = ∑ n·Keqn− 1·[M ]n n=1

n=1

=

by means of size exclusion chromatography according to eq 6. The observed linear relationship confirms the presence of the equilibrium between monomer and fibrils. The evaluated equilibrium constants are, respectively, 247.4 M−1, 628.4 M−1, 1575.7 M−1 and 9639.4 M−1 at pH 2, 2.5, 3.0, and 4.0. As shown in Figure 3b, the equilibrium constant increases together with the increase of the environmental pH and the decrease of the peptide net charge. These results indicate that the equilibrium is shifted toward fibrils at lower net charges, whereas the monomeric form is energetically favorable at higher net charges. The Gibbs free energy is related to the equilibrium constant by the equation

[M ] (1 − Keq ·[M ])2 (3) −2

which can be derived based on the series for (1 − x) . The equilibrium constant can be therefore defined as73 Keq =

1 − [M ]

1 [M ]·C tot

(4)

Replacing monomer concentration with monomer fraction, xM = [M]/Ctot, where Ctot is the total peptide concentration, the equilibrium constant is equal to 1 1 − xM = · C tot xM

Keq

ΔG = −R ·T ·ln(Keq)

where R is the gas constant and T is the temperature. The electrostatic contribution of the Gibbs free energy between two charged objects can be approximated as30,74

(5)

From eq 5, which can be rewritten as 1− xM

xM

= Keq ·C tot

(7)

ΔGe =

(6)

(z p·e)2 4·π ·ε0 ·εr ·r

·exp( −κ ·r )

(8)

where zp is the net charge of the peptide, e is the elementary charge, ε0 is the vacuum permittivity, εr is the relative dielectric permittivity, r is the distance between the ions, and κ is the inverse Debye length.

it can be seen that the monomer fraction is directly related to the total peptide concentration via the equilibrium constant. In Figure 3a, we plot the monomer content as a function of the total peptide concentration for various pH values as measured 3796

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules

4c). Above 50 mM, where the screening effect is saturated, the monomer amount is independent of the salt concentration. The charge screening effect is confirmed by the calculation of the change in the Debye length as a function of salt concentration (inset in Figure 4b), which shows a decrease in the Debye length with increasing salt concentration. This decrease in the Debye length is significant in particular at low ionic strengths, while the dependence starts to flatten out at salt concentrations larger than about 50 mM, suggesting that the most of the charges on the peptide surface are screened above this salt concentration. Perturbation of the Monomer−Fibril Equilibrium by Chemical Denaturation. We further investigated the thermodynamic stability of the fibrils upon chemical denaturation at the two reference pH values of 2.0 and 4.5, where, respectively, the smallest and the largest amount of fibrils were observed. We monitored the monomer content in 10 mM HCl (pH 2.0) and 10 mM acetate buffer (pH 4.5) with guanidinium hydrochloride (Gnd-HCl) at concentrations ranging from 0 to 6 M. The pH value was measured after the addition of the peptide and Gnd-HCl. As shown in Figure 5a, at both pH values we observe a nonmonotonic behavior of the monomer amount as a function of the Gnd-HCl concentration: the monomer percentage initially decreases as small amounts of Gnd-HCl are added into the system. A further increase in the concentration of Gnd-HCl destabilizes the fibrils and promotes the dissociation of monomers into the solution. Interestingly, at Gnd-HCl concentration of 6 M, the monomer content is similar at both pH 2.0 and 4.5 (41.9 ± 2.3%) (Figure 5a), despite that the stability of the fibrils in the absence of Gnd-HCl is different at the two pH values. Analysis by TEM imaging confirms that aggregates with fibrillar morphology are produced at all concentrations of denaturant (Figure S7a−d in the Supporting Information). Moreover, the disruption of β-sheet structures at 6 M Gnd-HCl is confirmed by CD analysis (Figure S7e in the Supporting Information). In addition to a charge screening effect due to the ionic nature of Gnd-HCl, the initial decrease of the monomer content observed after the addition of a low amount of GndHCl (0.3−1 M) can be correlated to a variation in the peptide conformation. To confirm this hypothesis, we performed metadynamics simulations of the structure of monomeric peptides at a reference concentration of denaturant equal to 0.6 M at both pH values under consideration. At pH 2.0, we observed an increase in the radius of gyration of the structure with the minimum energy from about 0.56 to 0.62 nm, as seen in Figure 5b (the detailed free energy surface is shown in Figure S8 in the Supporting Information). The global minimum of the free energy surface is still represented by random coil structures, which now exhibit a more open configuration due to the denaturing effect of guanidinium. The region of the free energy surface around the minimum (ΔF ≤ 2 kJ/mol) includes structures characterized by a wide span of Rg values (0.58−0.76 nm). This result highlights that, although some compact structures are still present, guanidinium shifts the equilibrium toward unfolded configurations of the monomer, which can enhance the propensity of the peptide toward aggregation, as observed by Zhang et al.70 Regions of the free energy surface that are more distant from the minimum (ΔF ≤ 4 kJ/mol) contain a substantial presence of unfolded structures (Figure S8 in the Supporting Information).

According to eq 8, if the electrostatic contribution dominates in the Gibbs free energy, a linear relationship between the free energy and the square of the peptide net charge is expected. Indeed, we observed a linear correlation between these two values (R2 value equal to 0.985) as seen in Figure 4a, suggesting that electrostatics plays an important role in the peptide selfassembly process.

Figure 4. Electrostatics in RADA 16-I self-assembly. (a) Gibbs free energy as a function of the net charge, zp, of RADA 16-I peptide. (b,c) Effect of salt on (b) the monomer fraction and (c) the equilibrium constant calculated from eq 5 for RADA 16-I solutions at 1 g/L in 10 mM HCl at pH 2.0 (●, left y-axis) and in 10 mM acetate buffer at pH 4.5 (■, right y-axis). The inset in (b) shows the change in the Debye length as a function of salt concentration. If error bars are not visible, they are smaller than the symbol. Representative SEC chromatograms are shown in Figure S5 in the Supporting Information.

In order to further investigate this effect, we measured the monomer content at both pH 2.0 and pH 4.5 as a function of the salt concentration. In the presence of electrostatic interactions, the addition of salt at a concentration of 10−50 mM screens the repulsion forces, thus promoting aggregation. Indeed, as shown in Figure 4b, an initial increase of the salt concentration induces a decrease of the monomer content and a corresponding increase of the equilibrium constant (Figure 3797

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules

bind guanidinium ions through hydrogen bonds. The interactions between aspartic acid residues and positively charged guanidinium ions are indeed more favorable at pH 4.5 than at pH 2.0, due to the unprotonated state of carboxyl groups. In order to further investigate the role of guanidinium hydrochloride on the aggregation propensity at pH 4.5, we performed a 60 ns molecular dynamics simulation at 300 K and 1 atm by placing two partially unfolded monomers in a 0.6 M denaturant solution: the molecular trajectories showed that the guanidinium ions bind to the dissociated aspartic acid side chains, thereby acting as salt bridge between peptides (Figure 5d) and promoting aggregation. Although this simulation cannot clearly provide a comprehensive picture of the overall mechanism of the formation of fibrils, this result highlights a possible role of the guanidinium hydrochloride in the first stages of the self-assembly process. The additional increase in the denaturant content from about 1 to 6 M causes further denaturation of the peptide, which leads to the dissociation of the monomer from the fibrils into the solution. Overall, these results indicate that, in addition to electrostatic effects, the monomer conformation and the charge distribution plays a significant role in the fibril formation, and that extended conformations favor the self-assembly process of the RADA 16-I peptide.



CONCLUSIONS We investigated the role of electrostatic interactions in the selfassembly of the model amphiphilic peptide RADA 16-I by using a combination of experimental characterization and molecular dynamic simulations. It is found that the optimal net charge to promote the formation of regular amyloid fibrils corresponds to a value of +0.75, a result that is in agreement with previous findings reported in the literature,37,72 indicating the possible presence of a certain level of universality in the aggregation behavior of peptides. Larger net charges, corresponding to low pH values, destabilize the fibrillar aggregates and promote the release of monomers that are in equilibrium with the fibrils, while lower net charge induces the formation of amorphous precipitates. The quantification of the fibril−monomer equilibrium constant as a function of pH and peptide net charge shows that electrostatic interactions contribute largely to the dependence of the free energy of fibril formation on the pH value. The addition of both salt and a charged destabilizer (guanidinium hydrochloride) at moderate concentration (0.3−1 M) at both pH values shifts the monomer-fibril equilibrium toward the fibrillar aggregates. The effect of salt can be easily explained by the screening of electrostatic repulsion. By contrast, the promotion of fibril formation in the presence of guanidinium hydrochloride is attributed to modifications of the peptide conformation in addition to charge screening effects. The experimental and modeling results of this work indicate that intermolecular electrostatic interactions play a key role in the complex combination of the several intermolecular forces that contribute to fibril formation and stability. These additional interactions include largely hydrogen bonding19 as well as “steric zippers” effects23 and bending energies related to the twisting of the fibrils.24−27 Despite the complexity of this behavior, which is reflected by the role of the peptide conformation and the surface charge distribution on the aggregation stability, the peptide net charge appears to be a

Figure 5. Chemical stability of RADA 16-I fibrils. (a) Amount of monomer in RADA 16-I solutions at 0.5 g/L in 10 mM HCl at pH 2.0 (⧫) and 10 mM acetate buffer at pH 4.5 (■) as a function of Gnd-HCl concentration in the range 0−6 M. If error bars are not visible, they are smaller than the symbol. (b−c) Free energy landscape as a function of the radius of gyration at pH 2.0 (b) and pH 4.5 (c). (d) Simulation illustrating the formation of a Gnd-HCl salt bridge between two interacting monomers in 0.6 M Gnd-HCl solution at pH 4.5. Representative SEC chromatograms are shown in Figure S6 in the Supporting Information.

At pH 4.5, the addition of 0.6 M denaturant induces a change in the radius of gyration from 0.66 to 0.62 nm as shown in Figure 5c. This change corresponds to the disruption of the βhairpin-like geometry into disordered and partially unfolded structures in the regions of the energy diagram close to the minimum (ΔF ≤ 6 kJ/mol). The partially unfolded conformation corresponding to the global minimum in the free energy surface in the presence of guanidinium exposes a larger surface area to the solvent (1718 Å2) with respect to the β-hairpin-like arrangement of the corresponding structure in pure water (1574 Å2). This change is likely to be driven by the exposure of dissociated aspartic acid residues, which can easily 3798

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules

(8) D'Auria, G.; Vacatello, M.; Falcigno, L.; Paduano, L.; Mangiapia, G.; Calvanese, L.; Gambaretto, R.; Dettin, M.; Paolillo, L. Selfassembling properties of ionic-complementary peptides. J. Pept. Sci. 2009, 15 (3), 210−219. (9) Zhang, S.; Greenfield, M. A.; Mata, A.; Palmer, L. C.; Bitton, R.; Mantei, J. R.; Aparicio, C.; de la Cruz, M. O.; Stupp, S. I. A selfassembly pathway to aligned monodomain gels. Nat. Mater. 2010, 9 (7), 594−601. (10) Adler-Abramovich, L.; Gazit, E. The physical properties of supramolecular peptide assemblies: from building block association to technological applications. Chem. Soc. Rev. 2014, 43 (20), 6881−6893. (11) Zhao, X.; Zhang, S. Designer Self-Assembling Peptide Materials. Macromol. Biosci. 2007, 7 (1), 13−22. (12) Holmes, T. C.; de Lacalle, S.; Su, X.; Liu, G.; Rich, A.; Zhang, S. Extensive neurite outgrowth and active synapse formation on selfassembling peptide scaffolds. Proc. Natl. Acad. Sci. U. S. A. 2000, 97 (12), 6728−6733. (13) Kisiday, J.; Jin, M.; Kurz, B.; Hung, H.; Semino, C.; Zhang, S.; Grodzinsky, A. J. Self-assembling peptide hydrogel fosters chondrocyte extracellular matrix production and cell division: Implications for cartilage tissue repair. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (15), 9996−10001. (14) Matson, J. B.; Newcomb, C. J.; Bitton, R.; Stupp, S. I. Nanostructure-templated control of drug release from peptide amphiphile nanofiber gels. Soft Matter 2012, 8 (13), 3586−3595. (15) Fung, S. Y.; Yang, H.; Chen, P. Sequence Effect of SelfAssembling Peptides on the Complexation and In Vitro Delivery of the Hydrophobic Anticancer Drug Ellipticine. PLoS One 2008, 3 (4), e1956. (16) Perale, G.; Rossi, F.; Santoro, M.; Peviani, M.; Papa, S.; Llupi, D.; Torriani, P.; Micotti, E.; Previdi, S.; Cervo, L.; Sundström, E.; Boccaccini, A. R.; Masi, M.; Forloni, G.; Veglianese, P. Multiple drug delivery hydrogel system for spinal cord injury repair strategies. J. Controlled Release 2012, 159 (2), 271−280. (17) Greenwald, J.; Riek, R. Biology of Amyloid: Structure, Function, and Regulation. Structure 2010, 18 (10), 1244−1260. (18) Hamley, I. W. Peptide Fibrillization. Angew. Chem., Int. Ed. 2007, 46 (43), 8128−8147. (19) Knowles, T. P. J.; Vendruscolo, M.; Dobson, C. M. The amyloid state and its association with protein misfolding diseases. Nat. Rev. Mol. Cell Biol. 2014, 15 (6), 384−396. (20) Alexandrescu, A. T. Amyloid accomplices and enforcers. Protein Sci. 2005, 14 (1), 1−12. (21) Gazit, E. Mechanisms of amyloid fibril self-assembly and inhibition. FEBS J. 2005, 272 (23), 5971−5978. (22) Krysmann, M. J.; Castelletto, V.; Kelarakis, A.; Hamley, I. W.; Hule, R. A.; Pochan, D. J. Self-Assembly and Hydrogelation of an Amyloid Peptide Fragment. Biochemistry 2008, 47 (16), 4597−4605. (23) Sawaya, M. R.; Sambashivan, S.; Nelson, R.; Ivanova, M. I.; Sievers, S. A.; Apostol, M. I.; Thompson, M. J.; Balbirnie, M.; Wiltzius, J. J. W.; McFarlane, H. T.; Madsen, A. O.; Riekel, C.; Eisenberg, D. Atomic structures of amyloid cross-[bgr] spines reveal varied steric zippers. Nature 2007, 447 (7143), 453−457. (24) Adamcik, J.; Jung, J.-M.; Flakowski, J.; De Los Rios, P.; Dietler, G.; Mezzenga, R. Understanding amyloid aggregation by statistical analysis of atomic force microscopy images. Nat. Nanotechnol. 2010, 5 (6), 423−428. (25) Knowles, T. P. J.; De Simone, A.; Fitzpatrick, A. W.; Baldwin, A.; Meehan, S.; Rajah, L.; Vendruscolo, M.; Welland, M. E.; Dobson, C. M.; Terentjev, E. M. Twisting Transition between Crystalline and Fibrillar Phases of Aggregated Peptides. Phys. Rev. Lett. 2012, 109 (15), 158101. (26) Aggeli, A.; Nyrkova, I. A.; Bell, M.; Harding, R.; Carrick, L.; McLeish, T. C. B.; Semenov, A. N.; Boden, N. Hierarchical selfassembly of chiral rod-like molecules as a model for peptide β-sheet tapes, ribbons, fibrils, and fibers. Proc. Natl. Acad. Sci. U. S. A. 2001, 98 (21), 11857−11862. (27) Adamcik, J.; Mezzenga, R. Adjustable twisting periodic pitch of amyloid fibrils. Soft Matter 2011, 7 (11), 5437−5443.

good indicator of the propensity of peptides toward aggregation.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b01092. Mass spectrometry analysis of monomeric RADA 16-I. Convergence of the free energy as a function of simulation time per replica for RADA 16-I under different experimental conditions. Representative SEC chromatograms of samples under the investigated conditions. The effect of the addition of Gnd-HCl on the structure of RADA 16-I peptide. Free energy surface maps for RADA 16-I in the presence of 0.6 M Gnd-HCl at pH 2.0 and 4.5. (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Address †

Department Chemistry, University of Cambridge, Cambridge, United Kingdom Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank Swiss National Science Foundation (Grant 200020-147137/1) for financial support. Moreover, Lonza (Visp, Switzerland), in particular Dr. Alessandro Butte’ and Dr. Francesca Quattrini, is gratefully acknowledged for kind donation of the material. The authors are also grateful to Dr. Matteo Salvalaglio for stimulating and fruitful discussions about metadynamics simulations. Moreover, the authors acknowledge the computational resources provided with Brutus cluster by ETH Zurich and the MS Service of ETH Zurich for the help with the MS analysis of the peptide. Paolo Arosio acknowledges the Marie Curie fellowship scheme for career development.



REFERENCES

(1) Chiti, F.; Dobson, C. M. Protein Misfolding, Functional Amyloid, and Human Disease. Annu. Rev. Biochem. 2006, 75 (1), 333−366. (2) Dobson, C. M. Protein misfolding, evolution and disease. Trends Biochem. Sci. 1999, 24 (9), 329−332. (3) Maji, S. K.; Perrin, M. H.; Sawaya, M. R.; Jessberger, S.; Vadodaria, K.; Rissman, R. A.; Singru, P. S.; Nilsson, K. P. R.; Simon, R.; Schubert, D.; Eisenberg, D.; Rivier, J.; Sawchenko, P.; Vale, W.; Riek, R. Functional Amyloids As Natural Storage of Peptide Hormones in Pituitary Secretory Granules. Science 2009, 325 (5938), 328−332. (4) Maji, S. K.; Schubert, D.; Rivier, C.; Lee, S.; Rivier, J. E.; Riek, R. Amyloid as a Depot for the Formulation of Long-Acting Drugs. PLoS Biol. 2008, 6 (2), e17. (5) Zhao, X.; Pan, F.; Xu, H.; Yaseen, M.; Shan, H.; Hauser, C. A. E.; Zhang, S.; Lu, J. R. Molecular self-assembly and applications of designer peptide amphiphiles. Chem. Soc. Rev. 2010, 39 (9), 3480− 3498. (6) Chen, P. Self-assembly of ionic-complementary peptides: a physicochemical viewpoint. Colloids Surf., A 2005, 261 (1−3), 3−24. (7) Cavalli, S.; Albericio, F.; Kros, A. Amphiphilic peptides and their cross-disciplinary role as building blocks for nanoscience. Chem. Soc. Rev. 2010, 39 (1), 241−263. 3799

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules (28) Leon, L.; Logrippo, P.; Tu, R. Self-Assembly of Rationally Designed Peptides under Two-Dimensional Confinement. Biophys. J. 2010, 99 (9), 2888−2895. (29) Grant, M. L. Nonuniform Charge Effects in Protein−Protein Interactions. J. Phys. Chem. B 2001, 105 (14), 2858−2863. (30) Buell, A. K.; Hung, P.; Salvatella, X.; Welland, M. E.; Dobson, C. M.; Knowles, T. P. J. Electrostatic Effects in Filamentous Protein Aggregation. Biophys. J. 2013, 104 (5), 1116−1126. (31) Majhi, P. R.; Ganta, R. R.; Vanam, R. P.; Seyrek, E.; Giger, K.; Dubin, P. L. Electrostatically Driven Protein Aggregation: βLactoglobulin at Low Ionic Strength. Langmuir 2006, 22 (22), 9150−9159. (32) Yan, Y.; Seeman, D.; Zheng, B.; Kizilay, E.; Xu, Y.; Dubin, P. L. pH-Dependent Aggregation and Disaggregation of Native βLactoglobulin in Low Salt. Langmuir 2013, 29 (14), 4584−4593. (33) Piazza, R. Interactions and phase transitions in protein solutions. Curr. Opin. Colloid Interface Sci. 2000, 5 (1−2), 38−43. (34) Piazza, R. Protein interactions and association: an open challenge for colloid science. Curr. Opin. Colloid Interface Sci. 2004, 8 (6), 515−522. (35) Zou, D.; Tie, Z.; Lu, C.; Qin, M.; Lu, X.; Wang, M.; Wang, W.; Chen, P. Effects of hydrophobicity and anions on self-assembly of the peptide EMK16-II. Biopolymers 2010, 93 (4), 318−329. (36) Zou, D.; Cao, Y.; Qin, M.; Dai, W.; Wang, W. Formation of [small alpha]-helix-based twisted ribbon-like fibrils from ioniccomplementary peptides. Chem. Commun. 2011, 47 (26), 7413−7415. (37) Aggeli, A.; Bell, M.; Carrick, L. M.; Fishwick, C. W. G.; Harding, R.; Mawer, P. J.; Radford, S. E.; Strong, A. E.; Boden, N. pH as a Trigger of Peptide β-Sheet Self-Assembly and Reversible Switching between Nematic and Isotropic Phases. J. Am. Chem. Soc. 2003, 125 (32), 9619−9628. (38) Mezzenga, R.; Fischer, P. The self-assembly, aggregation and phase transitions of food protein systems in one, two and three dimensions. Rep. Prog. Phys. 2013, 76 (4), 046601. (39) Krebs, M. R. H.; Devlin, G. L.; Donald, A. M. Protein Particulates: Another Generic Form of Protein Aggregation? Biophys. J. 2007, 92 (4), 1336−1342. (40) Barducci, A.; Bonomi, M.; Prakash, M. K.; Parrinello, M. Freeenergy landscape of protein oligomerization from atomistic simulations. Proc. Natl. Acad. Sci. U. S. A. 2013, 110 (49), E4708− E4713. (41) Bussi, G.; Gervasio, F. L.; Laio, A.; Parrinello, M. Free-energy landscape for beta hairpin folding from combined parallel tempering and metadynamics. J. Am. Chem. Soc. 2006, 128 (41), 13435−13441. (42) Saladino, G.; Pieraccini, S.; Rendine, S.; Recca, T.; Francescato, P.; Speranza, G.; Sironi, M. Metadynamics Study of a β-Hairpin Stability in Mixed Solvents. J. Am. Chem. Soc. 2011, 133 (9), 2897− 2903. (43) Berteotti, A.; Barducci, A.; Parrinello, M. Effect of Urea on the beta-Hairpin Conformational Ensemble and Protein Denaturation Mechanism. J. Am. Chem. Soc. 2011, 133 (43), 17200−17206. (44) Arosio, P.; Owczarz, M.; Wu, H.; Butté, A.; Morbidelli, M. Endto-End Self-Assembly of RADA 16-I Nanofibrils in Aqueous Solutions. Biophys. J. 2012, 102 (7), 1617−1626. (45) Becke, A. D. Density functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98 (7), 5648−5652. (46) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37 (2), 785−789. (47) Ditchfield, R.; Hehre, W. J.; Pople, J. A. Self Consistent Molecular Orbital Methods. IX. An Extended Gaussian Type Basis for Molecular Orbital Studies of Organic Molecules. J. Chem. Phys. 1971, 54 (2), 724−728. (48) Tomasi, J.; Mennucci, B.; Cammi, R. Quantum mechanical continuum solvation models. Chem. Rev. 2005, 105 (8), 2999−3093. (49) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Kollmann, P. A. Application of Resp Charges to Calculate Conformational Energies, Hydrogen-Bond Energies, and Free-Energies of Solvation. J. Am. Chem. Soc. 1993, 115 (21), 9620−9631.

(50) Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A WellBehaved Electrostatic Potential Based Method Using Charge Restraints for Deriving Atomic Charges - the Resp Model. J. Phys. Chem. 1993, 97 (40), 10269−10280. (51) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. v. R. Efficient diffuse function-augmented basis sets for anion calculations. III. The 3-21+G basis set for first-row elements. J. Comput. Chem. 1983, 4 (3), 294−301. (52) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.1; Gaussian, Inc.: Wallingford, CT, 2009. (53) Deighan, M.; Bonomi, M.; Pfaendtner, J. Efficient Simulation of Explicitly Solvated Proteins in the Well-Tempered Ensemble. J. Chem. Theory Comput. 2012, 8 (7), 2189−2192. (54) Laio, A.; Parrinello, M. Escaping free-energy minima. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (20), 12562−12566. (55) Barducci, A.; Bussi, G.; Parrinello, M. Well-tempered metadynamics: A smoothly converging and tunable free-energy method. Phys. Rev. Lett. 2008, 100 (2), 020603. (56) Pronk, S.; Pall, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.; Apostolov, R.; Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel, D.; Hess, B.; Lindahl, E. GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics 2013, 29 (7), 845−854. (57) Tribello, G. A.; Bonomi, M.; Branduardi, D.; Camilloni, C.; Bussi, G. PLUMED 2: New feathers for an old bird. Comput. Phys. Commun. 2014, 185 (2), 604−613. (58) Case, D. A.; Babin, V.; Berryman, J. T.; Betz, R. M.; Cai, D. S.; Cerutti, T. E.; Cheatham, T. E., III; Darden, T. A.; Duke, R. E.; Gohlke, H.; Goetz, A. W.; Gusarov, S.; Homeyer, N.; Janowski, P.; Kaus, J.; Kolossvàry, I.; Kovalenko, A.; Lee, T. S.; LeGrand, S.; Luchko, T.; Luo, R.; Madej, B.; Merz, K. M.; Paesani, F.; Roe, D. R.; Roitberg, A.; Sagui, C.; Salomon-Ferrer, R.; Seabra, G.; Simmerling, C. L.; Smith, W.; Swails, J.; Walker, R. C.; Wang, J.; Wolf, R. M.; Wu, X.; Kollman, P. A. Amber 14; University of California: San Francisco, 2014. (59) Olsson, M. H. M.; Sondergaard, C. R.; Rostkowski, M.; Jensen, J. H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pK(a) Predictions. J. Chem. Theory Comput. 2011, 7 (2), 525−537. (60) Maier, J. A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; Hauser, K.; Simmerling, C. ff14SB: Improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput. 2015, 11, 3696−3713. (61) Wang, J. M.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development and testing of a general amber force field. J. Comput. Chem. 2004, 25 (9), 1157−1174. (62) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 1983, 79 (2), 926−936. (63) Joung, I. S.; Cheatham, T. E. Determination of alkali and halide monovalent ion parameters for use in explicitly solvated biomolecular simulations. J. Phys. Chem. B 2008, 112 (30), 9020−9041. (64) Bussi, G.; Donadio, D.; Parrinello, M. Canonical sampling through velocity rescaling. J. Chem. Phys. 2007, 126 (1), 014101. 3800

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801

Article

Biomacromolecules (65) Parrinello, M.; Rahman, A. Polymorphic Transitions in SingleCrystals - a New Molecular-Dynamics Method. J. Appl. Phys. 1981, 52 (12), 7182−7190. (66) Darden, T.; York, D.; Pedersen, L. Particle Mesh Ewald - an N.Log(N) Method for Ewald Sums in Large Systems. J. Chem. Phys. 1993, 98 (12), 10089−10092. (67) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M. LINCS: A linear constraint solver for molecular simulations. J. Comput. Chem. 1997, 18 (12), 1463−1472. (68) Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J. A. Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. U. S. A. 2001, 98 (18), 10037−10041. (69) Ye, Z.; Zhang, H.; Luo, H.; Wang, S.; Zhou, Q.; Du, X.; Tang, C.; Chen, L.; Liu, J.; Shi, Y.-K.; Zhang, E.-Y.; Ellis-Behnke, R.; Zhao, X. Temperature and pH effects on biophysical and morphological properties of self-assembling peptide RADA16-I. J. Pept. Sci. 2008, 14 (2), 152−162. (70) Zhang, H.; Luo, H.; Zhao, X. Mechanistic Study of SelfAssembling Peptide RADA16-I in Formation of Nanofibers and Hydrogels. J. Nanotechnol. Eng. Med. 2010, 1 (1), 011007−6. (71) Cameselle, J. C.; Meireles Ribeiro, J.; Sillero, A. Derivation and use of a formula to calculate the net charge of acid-base compounds. Its application to amino acids, proteins and nucleotides. Biochem. Educ. 1986, 14 (3), 131−136. (72) López de la Paz, M.; Goldie, K.; Zurdo, J.; Lacroix, E.; Dobson, C. M.; Hoenger, A.; Serrano, L. De novo designed peptide-based amyloid fibrils. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (25), 16052− 16057. (73) Baldwin, A. J.; Knowles, T. P. J.; Tartaglia, G. G.; Fitzpatrick, A. W.; Devlin, G. L.; Shammas, S. L.; Waudby, C. A.; Mossuto, M. F.; Meehan, S.; Gras, S. L.; Christodoulou, J.; Anthony-Cahill, S. J.; Barker, P. D.; Vendruscolo, M.; Dobson, C. M. Metastability of Native Proteins and the Phenomenon of Amyloid Formation. J. Am. Chem. Soc. 2011, 133 (36), 14160−14163. (74) Leckband, D.; Israelachvili, J. Intermolecular forces in biology. Q. Rev. Biophys. 2001, 34 (02), 105−267.

3801

DOI: 10.1021/acs.biomac.5b01092 Biomacromolecules 2015, 16, 3792−3801