You are on page 1of 7

J. Phys. Chem.

B 2008, 112, 441-447

441

Phase Diagram for Assembly of Biologically-Active Peptide Amphiphiles


Stefan Tsonchev,*,,| Krista L. Niece,, George C. Schatz, Mark A. Ratner, and Samuel I. Stupp,
Department of Chemistry and Center for Nanofabrication and Molecular Self-Assembly, and Department of Materials Science and Engineering, and Feinberg School of Medicine, Northwestern UniVersity, EVanston, Illinois 60208, and Department of Chemistry, Northeastern Illinois UniVersity, Chicago, Illinois 60625 ReceiVed: August 5, 2007

We construct a phase diagram for self-assembling biologically active peptide amphiphiles. The structure and stability of the assemblies are studied as a function of pH and salinity of the solution. The general features of the phase diagram are predicted based on theoretical modeling of the self-assembly process, as well as experimental data, and further experiments are performed to verify and ascertain the boundary locations of the diagram. Depending on solution conditions, the amphiphiles can form cylindrical or spherical micelles, intermediate structures between these, or may not assemble at all. We also demonstrate that changing conditions may result in phase transitions among these structures. This type of phase diagram could be useful in the design of certain supramolecular nanostructures by providing information on the necessary conditions to form them.

1. Introduction Peptide amphiphiles comprise a new class of self-assembling molecules1-4 that form materials with nanoscale structure intended for biological applications such as regenerative medicine. We have recently reported on peptide amphiphiles that under favorable conditions self-assemble from aqueous solutions into cylindrical nanostructures, which subsequently bundle to form networks and, therefore, macroscopic gels.1,2 The amphiphilic molecules are designed in such a way that after self-assembly the biologically active functional groups appear on or near the surface of the nanostructures, where they are accessible to the external environment for cell signaling. Changing the conditions under which self-assembly takes place has a dramatic impact on the final shape and properties of the resulting supramolecular nanostructure. The structure does not form if certain conditions are not met. Therefore, it is of crucial importance to determine the most favorable conditions and most efficient means for the formation of a given targeted structure. Also, in many cases the intention is to synthesize molecules that will self-assemble under physiological conditions, and therefore, one must be able to control the self-assembly process within a narrow range of parameters. There has been a large number of studies in molecular self-assembly modeling5-13 as well as experimental work in this area.14-35 We have investigated the properties of a particular peptide amphiphile belonging to the general class of self-assembling materials discussed here, intended for use in regenerative medicine.1,36-39 It was shown that at pH 31,40
Part of the James T. (Casey) Hynes Festschrift. * To whom correspondence should be addressed. E-mail address: stefan@chem.northwestern.edu. Department of Chemistry and Center for Nanofabrication and Molecular Self-Assembly, Northwestern University. Department of Materials Science and Engineering, and Feinberg School of Medicine, Northwestern University. | Northeastern Illinois University.

this amphiphile assembles into cylindrical nanofibers that can bundle to form a gel. The gel can in turn be used as a matrix for cell proliferation. Our analysis determined that the cylindrical shape of the structures is the result of strong directional electrostatic interactions between large dipoles in the headgroups of the amphiphilic molecules under the conditions of the experiment, as well as formation of a sheet due to hydrogen bonding between the peptide segments of the molecules.37-39 During these studies it became clear that changing the conditions of the experiment would lead to different charges in the headgroups and hence to very different interactions between the amphiphiles. And indeed, it was observed for negatively charged PA molecule1,40 that the cylindrical fibers did not form, or they disassembled, when the pH of the solution was above the 2-4 range. Since these amphiphiles are intended for physiological applications, it would be desirable for self-assembly to occur at physiological pH. It is, therefore, advantageous to manipulate the shape and stability of the assembly by varying environmental conditions, and in particular, to be able to induce self-assembly at a pH 7. These considerations suggested the current investigation into how the process of self-assembly is affected by changing environmental parameters. Our goal in this work is to build a pH/salinity phase diagram for the peptide amphiphile reported in ref 1 using both modeling and experimental characterization. In section 2, we present some theoretical considerations, that determine the general shape of the phase diagram. Section 3 shows results from Monte Carlo and molecular dynamics simulations, intended to verify the theoretical predictions and to give an idea about the stability of the self-assembled structures in the different regions of the diagram. In section 4, we provide results from experiments that were performed to verify the theoretical predictions and to provide a quantitative determination of the two-phase lines in the diagram. Section 5 contains the discussion of results and conclusions.

10.1021/jp076273z CCC: $40.75 2008 American Chemical Society Published on Web 12/19/2007

442 J. Phys. Chem. B, Vol. 112, No. 2, 2008

Tsonchev et al.

Figure 1. Chemical structure of the self-assembling peptide amphiphile studied in this work.

from the tip. After determining D in this way, it is shortened by the distances between each lolly pop axis and its surface, to measure the actual distance between the lolly pop surfaces. The measure of the range of the hydrophobic interaction, , typically varies between 1-2 nm; here we have taken to be 2 nm. The electrostatic interaction between each pair of charges is given by

Eel )

q1q2e-r/ 4 0 r

(2)

Figure 2. Model of interactions between the peptide amphiphiles for the pH range 2-4. The parallel arrangement of the molecules is preferable for low salt concentration due to the more favorable electrostatic interactions between the (red) headgroups.37 In the case of very strong screening, the structure on the right is expected to become more stable. The opposite charges on the finite dipoles in the hydrophilic headgroups are shown in black and white.

2. Theoretical Considerations In our earlier work,37 we concluded that at pH values between 2 and 4 the peptide amphiphile shown in Figure 1 becomes a zwitterion possessing a large dipole in its hydrophilic headgroup, perpendicular to the backbone of the molecule. We explained the cylindrical shape of its self-assemblies as a result of directional electrostatic interactions between the dipoles in the headgroups. A simple scheme, shown in Figure 2, where the interacting molecules are modeled in a coarse-grained manner as lolly pops, or chains of beads, explains the competition between the electrostatic and hydrophobic interactions. In this qualitative model, the dipoles could be also considered as including the hydrogen bonds between the molecules, as these also result from electrostatic interactions and have a well-defined direction. The hydrophobic interaction between the lolly pops by itself would lead to the formation of a spherical micelle36 and is modeled by an interaction potential of the form41

Eh ) -e-D/

(1)

where is a surface tension constant which is treated as a free parameter used to calibrate the equation, so that it gives the proper hydrophobic energy per particle in the hydrophobic core of the final assembly. In modeling the molecule in Figure 1, a value of of 0.7787 in units of kT at 298 K is appropriate. D is the distance between the tails of the lolly pops, which is determined in the following way. It is taken as the minimum average distance between two points on the lolly pop tail. The first point is the center of the smallest sphere in the tail, and the second is located on the lolly pop axis at 0.44 length units

where q1 and q2 are the two interacting charges separated by a distance r, 0 is the dielectric permittivity of free space, is the dielectric constant of the solution, taken as 78.5, and is the Debye length. The considerations which led to the explanation of the cylindrical structure are based on the higher stability of the parallel formation of lolly pops, assuming that the salt concentration in the solution is low, as in the actual experiment. However, it is clear that increased salt concentration would lead to screening of the charges, producing a weaker interaction between them. Generally speaking, at some point, the electrostatics would no longer dominate the hydrophobic forces, and the structure in the right of Figure 2 would be more stable. Thus, we could try to estimate when the electrostatic interaction would be offset by the hydrophobic stabilization. Naively, one could think that by using eqs 1 and 2 one could easily construct a transcendental equation for from which to estimate the salt concentration necessary to make the process to the right in Figure 2 preferable. However, the salt concentration calculated in this way is very high, a few molar, and therefore, this approach cannot be used, since eq 2 assumes the Debye-Huckel theory, which is valid for much lower concentrations, on the order of 10-2 M. Hence, we have relied on experiment to determine the level of screening necessary to induce a possible phase transition between the cylindrical and spherical structures in this pH range. On the basis of the pKa values of the component amino acids, we expect that, at a pH lower than 2, the amphiphile molecules would possess a charge of +1 each, due to the positive charge on the guanidino group of the arginine. In this case, assuming for simplicity that the charges are located at the centers of the largest spheres of the hydrophilic headgroups, it is easy to calculate the electrostatic repulsion for two particles at contact as shown in Figure 3. For f (infinite dilution), the electrostatic repulsion without the hydrogen bonding amounts to about 0.5 kT at 298 K, whereas the hydrophobic attraction between the particles stabilizes the system by -0.78 kT, compared to the reference state of the two particles infinitely far apart. The total stabilization of about -0.28 kT would not be enough to keep the particles together against their thermal motion if it were not for the hydrogen bonding between them.

Phase Diagram for Peptide Amphiphiles

J. Phys. Chem. B, Vol. 112, No. 2, 2008 443

Figure 3. At pH f 0, the two particles (now charged) would stay together in a spherical formation, provided the salt concentration in the solution is high enough to screen the electrostatic repulsion between the charges. The charges are shown as black dots in the centers of the largest spheres of the hydrophilic headgroups.

Figure 5. When the pH is lowered from 3 to about 0 while eliminating the charges in the headgroups or making them all positive and screened, the system starts as a cylinder, which collapses into a spherical micelle after an equilibration period. The initial cylindrical configuration is shown on the left, and a sample snapshot of the equilibrium spherical configuration is shown on the right.

Figure 4. Left panel shows a sample of an equilibrium configuration of a system of 194 particles with a charge of +1 each in a solution of 0.016 M monovalent salt concentration. In a solution of 0.023 M monovalent salt concentration, the particles assemble into a spherical micelle, shown on the right. The structure is dynamic in nature, with particles constantly leaving and entering it, while preserving its overall spherical shape.

This was found in our earlier work39 to be about 2.7 H-bonds per particle, which is more than 10 kT. Hence, we would expect the particles to stay together in a cylindrical assembly until a pH of about or below 0, when protonation of the backbone amide oxygens in the cysteines would destroy the intermolecular hydrogen bonding and the structures would disassemble, as has been observed in the experiments.40 At this pH, increasing the salt concentration in the solution to a sufficiently high value would screen the charges and allow the particles to assemble into spheres, since the hydrogen bonding, which favors the formation of cylindrical structures, would be absent. In the next section, we will demonstrate this point with Monte Carlo simulations. Increasing the pH above 4 to about 7 could lead to a total charge of -3 in the hydrophilic headgroups due to the negative charges in the phosphate and carboxylic groups and the positive charge on the arginine. When the pH reaches about 9, the thiol groups in the cysteines, whose pKas are approximately 9, could also acquire a negative charge, bringing the net charge possibly up to as high as -7. After the pH exceeds about 12.5, the arginine would be neutralized and the total charge could become as high as -9. Thus, from the titration points (pKas) of the component amino acids (see the marked inflection points in Figure 7), we could infer transition lines of a phase diagram and predict possible phase transitions along the pH coordinate. For negligible salt concentration, in the pH range 4-9, the

electrostatic repulsion is calculated to be between 2.0 and 4.0 kT ,which would overcome the hydrophobic attraction between the tails; however, the hydrogen bonding is still expected to be stronger, so we would expect mainly cylindrical nanostructures in this range, albeit not so stable as the ones observed for the pH range 2-4. Relatively low salt concentration should be enough to stabilize the cylindrical structures and allow them to form the fibers necessary for the gel formation. This is the targeted range around neutral pH for the envisioned physiological applications of these amphiphiles. For pH > 9 and no added salt, a simple calculation shows that the electrostatic repulsion drastically increases up to about 18 kT and cannot be compensated by the hydrophobic interaction and the hydrogen bonding in the -sheet. Using eq 2 to predict the salt concentration needed to screen the electrostatic repulsion to the level of these attractive interactions is inadequate, since the obtained salt concentrations are far above the 10-2 M limit of the Debye-Huckel theory. Here, once again, we will rely on experiment to find the transition lines along the salinity coordinate of the diagram, where the screening of the charges is strong enough to allow the formation of the cylindrical nanostructures. We will also perform force field molecular dynamics simulations on the system in the pH range 9-12.5 at zero added salt and in 5 M NaCl solution, to demonstrate the stability of the structure at these concentrations and to illustrate the screening effect of the ions. 3. Phase Transitions Predicted from Computer Simulations 3.1. Monte Carlo Simulations in the Limit pH f 0. We will use the approach developed in our earlier work37 to perform Monte Carlo simulations on a system consisting of a large number of amphiphiles modeled as lolly pops with a charge of +1 at the centers of their hydrophilic headgroups, which would be the charge of the amphiphiles in the limit pH f 0, and hydrophobically interacting tails. We start the simulations by placing N particles in a cubic box of length L ) 30.0, in relative units for which the distance between the center of the largest sphere and the tip of each lolly pop is unity. The particles are distributed at random, except for a fraction which are put in the center of the box, close to each other, to serve potentially as a nucleation kernel for the micelle that might be formed. The particles interact according to eqs 1 and 2 as explained earlier and are moved one at a time. The motion is somewhat biased along the length of the particle to speed up the simulation in the case when a micelle is formed.

444 J. Phys. Chem. B, Vol. 112, No. 2, 2008

Tsonchev et al.

Figure 6. Equilibrium structure at pH ) 3 and zero salt concentration is used as a starting configuration for the simulations at pH ) 11 and shown in the left panel. The P atoms are shown in red, the S atoms are in green, the C atoms of the guanidino group of the arginine are in blue, and the rest of the atoms are in gray. The dipoles near the surface, determined by the negatively charged phosphate (red) and positively charged guanidino group of the arginine (blue), and the hydrogen bonds between the cysteines, which are outlined by the green S atoms and form a -sheet in the lower part of the headgroups, contribute to the order along one of the periodic axes, which is identified as the cylindrical axis of the full nanostructure.38,39 In the central panel is shown a sample of the resulting structure after several nanoseconds of molecular dynamics simulations at pH ) 11 with no added salt. The destruction of both the dipole-dipole and hydrogen-bonding order is clearly visible. In this structure the head and tail regions of some of the peptide amphiphiles have gotten mixed together, while in the structures shown in the other panels heads interact with heads, and tails with tails. In the right panel we show the structure resulting from the simulations at pH ) 11 when an additional 94 Na+ and Clions are added to the system, approximately a 5 M solution. In this case the original pH 3 structure remains intact due to the screening effect of the added salt. In all panels the Na+ and Cl- ions, hydrogen atoms, and water molecules are not shown, for simplicity.

Figure 7. pH/salinity phase diagram of the self-assembling peptide amphiphiles. The experimental verification points (derived from a combination of visual examination, rheology, TEM) are shown with the figures containing the letters g, i, s, and n, as explained above. Under the diagram, we have shown the titration curve of the amphiphile, where the titration inflection points corresponding to the transition lines of the diagram are shown. The inset shows the dependence of the damping factor G/G on pH, with G/G < 0.2 corresponding to a gel. This provides us with a quantitative picture of the degree of gelation, which is correlated with the length and stability of the fibers.

To investigate the dependence of the resulting structure on salinity, we have varied , corresponding to different salt concentrations. At ) 24 , corresponding to a 0.016 M monovalent salt solution, we observe no assembly of particles. Instead, after an equilibration period, the initial kernel falls apart and the particles scatter in the box and remain in such a state indefinitely, as can be seen from the left panel of Figure 4. However, reducing to 20 , which is equivalent to increasing the salt concentration to about 0.023 M, has a dramatic effect

on the system. As a result of the stronger screening, now the hydrophobic interaction prevails and keeps the particles together. After an initial equilibration period, the kernel grows to include almost all of the particles, which form a dynamical spherical micelle, as shown in the right panel of Figure 4. The limit pH f 0 was achieved experimentally by placing the self-assembled peptide amphiphile nanofibers in HCl vapor,40 and in the above simulations, we have only taken into account the charges on the amphiphiles and the dissolved salt,

Phase Diagram for Peptide Amphiphiles whose concentrations are reasonably small so that the DebyeHuckel approach used here is valid. However, it is clear that the calculations are only qualitatively useful, and in a true solution of pH f 0, there would be a much higher concentration of charged species, which would invalidate the Debye-Huckel approach. The simulations predict that a phase transition between no assembly and a spherical micelle would occur if the salt concentration is increased past about 0.02 M for pH f 0. Consequently, we would expect that another phase transition, between a cylindrical and a spherical micelle, would be observed at a salt concentration higher than 0.02 M when lowering the pH from the range 2-4 to near 0. This can be demonstrated by another Monte Carlo simulation which starts from a cylindrical micelle, with the charges in the headgroups of the amphiphiles switched to positive and screened, or eliminated for simplicity, as shown in the left panel of Figure 5. It is observed that after an initial period of equilibration the cylindrical micelle collapses into a spherical one and then remains in that state indefinitely. Figure 5 shows the initial state and a snapshot sample of the final configuration. Unlike this transition, the micelle remains in the cylindrical configuration indefinitely if the conditions are not changed, i.e., pH ) 3.37 3.2. Molecular Dynamics Simulations in the pH Range 9-12.5. In this pH range, the amphiphiles are highly charged. Here, as well as in the cases of high salinity, the DebyeHuckel theory is inadequate due to the high concentration of charged species in solution. Therefore, for these cases we have used the MM3 force field to perform all-atom molecular dynamics simulations,42 implemented in the Tinker package,43 as explained earlier.39 These simulations were run within the canonical ensemble using the Berendsen algorithm,44 and the hydrogen atoms were constrained at their ideal bond distance using the rattle algorithm.45 As before, we have performed the simulations in 2D periodic boundary conditions, starting from the equilibrium structure obtained by us earlier for pH ) 3,39 which was taken as the initial configuration for a molecular dynamics simulation at pH ) 11. For this purpose, we removed the hydrogens from the four thiols of the cysteines of each molecule, as well as from the carboxylic groups, and the corresponding sulfur and oxygen atoms were charged accordingly. To balance the charges, we added 96 Na+ ions in the elementary cell. Also, we have used explicit solvent, adding 450 water molecules above the surface of the assembly. This was enough to form at least three layers and ensure proper screening of the charges, after an equilibration period during which the water molecules soaked the hydrophilic headgroups. First, we performed simulations on the system with no added salt for several nanoseconds. The initial assembly and a sample of the resulting structure from these simulations are shown in Figure 6, in the left and central panels, respectively. It is obvious from Figure 6 that when no salt is added to the solution the electrostatic repulsion between the charged thiols and carboxylic groups destroys the -sheet of the self-assembly, as well as the ordering of the dipoles in the headgroups. Therefore, at pH ) 11 and with no added salt in the solution to screen the electrostatic repulsion, we expect that the amphiphiles would not assemble. On the other hand, if we add enough Na+ and Cl- ions, to make up about 5 M solution (in this case 94 ions each), the screening due to these ions is enough to prevent the micelle from disassembling, and the regularity of the -sheet and dipolar ordering in the headgroups is largely retained, as can be inferred from the right panel of Figure 6. Thus, it is clear that, even at much higher than physiological pH, the

J. Phys. Chem. B, Vol. 112, No. 2, 2008 445 amphiphiles would still assemble if enough salt is added in the solution to reduce their electrostatic repulsion. 4. Experimental Methods 4.1. General Information. Synthesis of the peptide amphiphile (PA) in Figure 1 was performed as described previously.1 All other reagents were purchased from Aldrich Chemical Co. (Milwaukee, WI) unless otherwise stated and used as received. Synthesis was confirmed by mass spectrometry and 1H NMR. Purity was estimated at 85% using analytical HPLC. The PA was dissolved in slightly basic water containing 40 molar equiv of dithiothreitol to reduce any disulfide bonds, then aliquoted, lyophilized, and stored desiccated at -20 C until needed. 4.2. pH Titration. PA was dissolved at 0.1% by weight in Millipore water and sufficient potassium hydroxide to bring the pH to 12.87 as determined by a Mettler-Toledo Delta 320 pH meter. Basicity was confirmed independently by pH paper. Dilute hydrochloric acid (HCl) was added in aliquots and the resulting pH recorded after stabilization. Approximate pKas were determined by plotting pH versus HCl added and identifying the corresponding inflection points. 4.3. Oscillating Rheometry. PA was dissolved at 1% by weight in Millipore water and mixed with various quantities of HCl on the stage of a Paar Physica Modular Compact Rheometer (MCR) 300. A 25 mm parallel plate configuration was used. After 10 min, storage (G) and loss (G) moduli were measured at 3% strain in a frequency sweep from 100 to 0.1 Hz. These parameters were as determined previously.23 The pH of the solution or gel was measured after the frequency sweep using pH paper. 4.22 Hz was chosen from the middle of the linear viscoelastic region of the sweep, and data at that frequency were used to determine the damping factor (G/G). For pH 5, 3.16 Hz was used instead, as the linear region ended at this frequency. At pH 6.5, no gelation occurred as gauged by the ratio of storage and loss moduli, and the damping factor was always greater than one. All data are the average of two runs. 4.4. Transmission Electron Microscopy. PA was dissolved at 1% by weight in Millipore water and then diluted 10 in aqueous solutions of various pH and sodium chloride (NaCl) salt concentration. The pH was measured and recorded after this dilution using pH paper. Visual observations with regard to gelation were noted at this point as well. After 1 h, 0.1 N iodine solution was added in 100 L aliquots until the solution remained brown. This was done to oxidize the thiol groups and covalently capture the solution structure. After several hours, samples containing salt at concentrations higher than 1 M were dialyzed against water at the same pH as the PA solution. After hours to days, TEM samples were prepared by drop-casting 10 L of this PA solution onto Quantifoil 1.2/2.3 carbon-coated copper grids. The residue was wicked off after 1 min and the grid rinsed once with distilled water before drying. Samples were stained using 2% phosphotungstic acid in water and viewed on an HF8100 Hitachi TEM. 5. Discussion and Conclusions From theoretical considerations and calculations as well as experimental results presented above, we have been able to build a semiquantitative pH/salinity phase diagram for the selfassembling peptide amphiphile studied here (shown in Figure 7). The phase transitions along the pH axis are determined by the equivalence points of the amphiphile, as confirmed by its titration curve shown in the figure. The transition lines along the salinity axis are determined by theory and experiment, and

446 J. Phys. Chem. B, Vol. 112, No. 2, 2008

Tsonchev et al.

Figure 8. TEM pictures of the self-assembling amphiphiles. On the left are shown fibers bundled together and forming a gelatinous structure at pH ) 0.5 and zero salinity. The right panel shows the disassembled system at pH ) 13.0 in a 0.9 M salt solution, where no structure is seen.

their agreement is evident in cases where a comparison has been possible. It should be pointed out that the actual transition lines are smeared and there is a continuous transition between the different phases, as has been indicated by the transition regions, where an intermediate state exists, with short fibers that are unable to form a gel. The difference between the various structures existing in the different regions of the phase diagram can be seen from the TEM pictures shown in Figure 8, where we see gel-forming fibers at low pH and low salinity, contrasted to the disassembled state of the highly charged amphiphiles at high pH and relatively low salinity. An important part of the phase diagram, from a practical perspective, is the region around pH ) 7 because it applies to physiological conditions. Here we see the advantage of having a phase diagram for such a system, by recognizing that it can suggest obvious means for self-assembling the amphiphiles in a physiological environment. Namely, increasing the salinity of the environment to below 1 M at pH 7, allows for selfassembly and gelation, which would be necessary for the formation of a gel matrix for in vivo applications. We have used theoretical and experimental tools to derive a pH/salinity phase digram of self-assembling peptide amphiphiles intended initially for use in regenerative medicine. The diagram shows the regions in which the amphiphiles would assemble into fibers, forming a gel that could be used as a matrix for cells. The gradual phase transitions between cylindrical fibers and spherical micelles, or a disassembled state, are also shown. We have shown how theory and experiment can be combined in systematic studies of a complex self-assembly process, to produce a phase diagram describing how shape and stability of a self-assembled nanostructure depend on environmental conditions. Acknowledgment. This paper is in honor of Casey Hynesbrilliant chemist, gifted teacher, visionary scientist and warm friend. This research was supported by grants from DoE (Grant No. DE-FG0200RT45810/A006), DoD/MURI (Grant No. F49620-00-1-0283), NIH (Grant No. SR01EB003806-02), NSFMRSEC (Grant No. DMR 0520513), and NSF (Grant No. CHE0550497). The authors also thank Professor Wes Burghardt at Northwestern University for access to the rheometer used in these experiments. The TEM work was performed in the EPIC facility of NUANCE Center at Northwestern University. NUANCE Center is supported by NSF-NSEC, NSF-MRSEC, Keck Foundation, the State of Illinois, and Northwestern University. Supporting Information Available: 1H NMR of PA, HPLC trace, and full rheology data. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes


(1) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Science 2001, 294, 1684. (2) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5133. (3) Yu, Y. C.; Berndt, P.; Tirrell, M.; Fields, G. B. J. Am. Chem. Soc. 1996, 118, 12515. (4) Yu, Y. C.; Roontga, V.; Daragan, V. A.; Mayo, K. H.; Tirrell, M.; Fields, G. B. Biochemistry 1999, 38, 1659. (5) Troisi, A.; Wong, V.; Ratner, M. A. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 255. (6) Wu, D.; Chandler, D.; Smith, B. J. Phys. Chem. 1992, 96, 4077. (7) Sinyagin, A.; Belov, A.; Kotov, A. Modelling Simul. Mater. Sci. Eng. 2005, 13, 389. (8) Mayer, B.; Kohler, G.; Rasmussen, S. Phys. ReV. E 1997, 55, 4489. (9) Shinoda, K.; Shinoda, W.; Liew, C. C.; Tsuzuki, S.; Morikawa, Y.; Mikami, M. Surf. Sci. 2004, 556, 109. (10) Lee, J. Y.; Shou, Z.; Balasz, A. C. Phys. ReV. Lett. 2003, 91, Art. No. 136103. (11) Aranovich, G. L.; Donohue, M. D. J. Chem. Phys. 2002, 116, 7255. (12) Johnson, R. A.; Nagaranjan, R. Colloids Surf., A 2000, 167, 21. Johnson, R. A.; Nagaranjan, R. Colloids Surf., A 2000, 167, 37. (13) Solis, F. J.; Stupp, S. I.; de la Cruz, M. O. J. Chem. Phys. 2005, 122, 054905. (14) Niece, K. L.; Hartgerink, J. D.; Donners, J. J. J. M.; Stupp, S. I. J. Am. Chem. Soc. 2003, 125, 7146. (15) Silva, G. A.; Czeisler, C.; Niece, K. L.; Beniash, E.; Harrington, D. A.; Kessler, J. A.; Stupp, S. I. Science 2004, 303, 1352. (16) Sone, E. D.; Stupp, S. I. J. Am. Chem. Soc. 2004, 126, 12756. (17) Arnold, M. S.; Guler, M. O.; Hersam, M. C.; Stupp, S. I. Langmuir 2005, 21, 4705. (18) Beniash, E.; Hartgerink, J. D.; Storrie, H.; Stupp, S. I. Acta Biomater. 2005, 1, 387. (19) Bull, S. R.; Guler, M. O.; Bras, R. E.; Meade, T. J.; Stupp, S. I. Nano Lett. 2005, 5, 1. (20) Guler, M. O.; Soukasene, S.; Hulvat, J. F.; Stupp, S. I. Nano Lett. 2005, 5, 249. (21) Guler, M. O.; Claussen, R. C.; Stupp, S. I. J. Mater. Chem. 2005, 15, 4507. (22) Jiang, H.; Stupp, S. I. Langmuir 2005, 21, 5242. (23) Stendahl, J. C.; Rao, M. S.; Guler, M. O.; Stupp, S. I. AdV. Funct. Mater. 2006, 16, 499. (24) Tovar, J. D.; Claussen, R. C.; Stupp, S. I. J. Am. Chem. Soc. 2005, 127, 7337. (25) Ozbas, B.; Kretsinger, J.; Rajagopal, K.; Schneider, J. P.; Pochan, D. J. Macromolecules 2004, 37, 7331. (26) Zhang, S. Nat. Biotechnol. 2003, 21, 1171. (27) Caplan, M. R.; Moore, P. N.; Zhang, S.; Kamm, R. D.; Lauffenburger, D. A. Biomacromolecules 2000, 1, 627. (28) Hartgerink, J. D.; Grania, J. R.; Milligan, R. A.; Ghadiri, M. R. J. Am. Chem. Soc. 1996, 118, 43. (29) Uversky, V. N.; Fink, A. L. BBA-Proteins Proteom. 2004, 1698, 131. (30) Dong, J.; Apkarian, R. P.; Lynn, D. G. Bioorg. Med. Chem. 2005, 13, 5213. (31) Iwaura, R.; Yoshida, K.; Masuda, M.; Yase, K.; Shimizu, T. Chem. Mater. 2002, 14, 3047. (32) Ma, Y.; Kolotuchin, S. V.; Zimmerman, S. C. J. Am. Chem. Soc. 2002, 124, 13757. (33) Watkins, D. M.; Sayed-Sweet, Y.; Klimash, J. W.; Turro, N. J.; Tomalia, D. A. Langmuir 1997, 13, 3136.

Phase Diagram for Peptide Amphiphiles


(34) Kato, T.; Kihara, H.; Ujiie, S.; Uryu, T.; Frechet, J. M. J. Macromolecules 1996, 29, 8734. (35) Hasenknopf, B.; Lehn, J.-M.; Kneisel, B. O.; Baum, G.; Fenske, D. Angew. Chem. Int. Ed. 1996, 35, 1838. (36) Tsonchev, S.; Schatz, G. C.; Ratner, M. A. Nano Lett. 2003, 3, 623. (37) Tsonchev, S.; Schatz, G. C.; Ratner, M. A. J. Phys. Chem. B 2004, 108, 8817. (38) Tsonchev, S.; Troisi, A.; Schatz, G. C.; Ratner, M. A. Nano Lett. 2004, 4, 427. (39) Tsonchev, S.; Troisi, A.; Schatz, G. C.; Ratner, M. A. J. Phys. Chem. B 2004, 108, 15278.

J. Phys. Chem. B, Vol. 112, No. 2, 2008 447


(40) Hartgerink, J. D. private communication. (41) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press: London, 1992. (42) Allinger, N. L.; Yuh, Y. H.; Lii, J.-H. J. Am. Chem. Soc. 1989, 111, 8551. Lii, J.-H.; Allinger, N. L. J. Comput. Chem. 1998, 19, 1001. (43) Ponder, J. W.; Richards, F. M. J. Comput. Chem. 1987, 8, 1016. Kundrot, C. E.; Ponder, J. W.; Richards, F. M. J. Comput. Chem. 1991, 12, 402. Dudek, M. J.; Ponder, J. W. J. Comput. Chem. 1995, 6, 791. (44) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. J. Chem. Phys. 1984, 81, 3684. (45) Andersen, H. C. J. Comput. Phys. 1983, 52, 24.

You might also like