Expanding the Nanoarchitectural Diversity Through Aromatic Di- and Tri-Peptide Coassembly: Nanostructures and Molecular Mechanisms Cong Guo,† Zohar A. Arnon,‡ Ruxi Qi,† Qingwen Zhang,*,§ Lihi Adler-Abramovich,∥ Ehud Gazit,*,‡ and Guanghong Wei*,† †
Key Laboratory for Computational Physical Sciences (MOE), State Key Laboratory of Surface Physics, and Department of Physics, Fudan University, Shanghai 200433, China ‡ Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences and ∥Department of Oral Biology, the Goldschleger School of Dental Medicine, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv 69978, Israel § College of Physical Education and Training, Shanghai University of Sport, 399 Changhai Road, Shanghai 200438, China S Supporting Information *
ABSTRACT: Molecular self-assembly is pivotal for the formation of ordered nanostructures, yet the structural diversity obtained by the use of a single type of building block is limited. Multicomponent coassembly, utilized to expand the architectural space, is principally based on empirical observations rather than rational design. Here we report large-scale molecular dynamics simulations of the coassembly of diphenylalanine (FF) and triphenylalanine (FFF) peptides at various mass ratios. Our simulations show that FF and FFF can co-organize into both canonical and noncanonical assemblies. Strikingly, toroid nanostructures, which were rarely observed for the extensively studied FF or FFF, are often seen in the FF-FFF coassembly simulations and later corroborated by scanning electron microscopy. Our simulations demonstrate a wide ratio-dependent variation of nanostructure morphologies including hollow and solid assemblies, much richer than those formed by each individual moiety. The hollow-solid structural transformation displays a discontinuous transition feature, and the toroids appear to be an obligatory intermediate for the structural transition. Interaction analysis reveals that the hollow-solid structural transition is mostly dominated by FF−FFF interactions, while the nanotoroid formation is determined by the competition between FF−water and FFF−water interactions. This study provides both structural and mechanistic insights into the coassembly of FF and FFF peptides, thus offering a molecular basis for the rational design of bionanomaterials utilizing peptide coassembly. KEYWORDS: diphenylalanine, triphenylalanine, controllable coassembly, nanostructural diversity, geometry map, coassembly mechanism, toroid nanostructure form nanovesicles,16−18 nanowires,19−21 nanoribbons,22 nanofibers,23 microrods,24 and hydrogel25 depending on different experimental conditions.26 These FF nanoassemblies have a broad range of applications including drug delivery, bioimaging, biosensors, and biological semiconductors.21,26−31 Different from their dipeptide counterpart, phenylalanine (F) peptides self-organize into amyloid fibrils32,33 linked with phenyl-
P
eptide nanostructures play an important role in bionanomaterial design due to their inherent biocompatibility and diverse applications in biology, nanomedicine, and nanotechnology.1−4 The combination of the 20 coded amino acid residues displays a vast variety of building blocks, which can lead to rich structural diversity.5−11 One of the extensively studied building blocks is the diphenylalanine (FF). FF peptides were originally found to form well-ordered nanotubes in aqueous solution,12 which can be used as nanoscale molds for the casting of metallic nanowires13 and can be aligned in strong magnetic fields.14,15 Thereafter, numerous studies revealed that FF-based peptides can also © 2016 American Chemical Society
Received: April 24, 2016 Accepted: August 22, 2016 Published: August 22, 2016 8316
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
www.acsnano.org
Article
ACS Nano Table 1. Summary of the MD Simulations for all of the FF-FFF Systemsa
a
There are 16 different systems with the FFF mass percentage (PFFF) ranging from 0 to 1. Multiple independent MD runs were performed for each system, and the simulation time of each MD run spans from 1.2 to 4.8 μs. The starting state of each MD run is a homogeneous mixture of FF and FFF peptides in water. We present in the last two columns the side view and the cross-section view of a representative snapshot of the most abundant nanostructures observed for each system. In each snapshot, the CG beads are colored as follows: FF main chain (red), FF side chain (white), FFF main chain (blue), and FFF side chain (yellow). Water molecules are not shown for clarity.
FF or FFF nanostructures are hydrogen bonding and interactions between aromatic moieties.26,34,38,40,43 Water− peptide interactions also influence the interpeptide interactions and the molecular packing in the FF or FFF assemblies.19,38,44,45 Although the specific and quantitative roles of these interactions still need to be qualified, their delicate balance results in the formation of different FF or FFF nanostructures.38,40 An interesting question is whether a mixture of FF and FFF peptides can form new nanostructures. If so, then what is the coassembly mechanism and how does one type of peptide (such as FFF) modulate the nanostructure formation of the other type of peptide? Can we achieve tunable peptide coassembly? To answer these questions, we systematically investigated the coassembly of FF and FFF peptides by performing 290 independent MD simulations of FF-FFF mixture at a FFF mass percentage (PFFF) ranging from 0 to 1 using coarse-grained MARTINI 2.1 force field.46,47 Our previous studies38,40 demonstrated that the simulation results,
ketonuria (a metabolic disorder disease), and triphenylalanine (FFF) peptides mainly self-assemble into solid nanoplates34 and nanospheres.35 Inspired by above-mentioned experimental works, increasing computational studies have focused on the molecular mechanism of F,36 FF,34,37−40 or FFF self-assembly.34,40 Interestingly, by using coarse-grained molecular dynamics (CG-MD) simulations, Tuttle, Ulijn, and co-workers explored the sequence space of di- and tripeptide self-assembly.41,42 Their simulations demonstrated that FW, WF, FF, and WW displayed the highest aggregation propensities, consistent with experiments.41 They also discovered that four FFF analogues, PFF, MFF, VFF, and FFM, are among the five highest aggregation propensity tripeptides (in addition to WFL) and FFF is the tripeptide with the seventh highest aggregation potential among the sequence space of 8000 tripeptides.42 Their work further demonstrates the central role of the FF motif and its extended FFF analogue as key elements in molecular self-assembly. The assembly driving forces of these 8317
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
Figure 1. Coassembly of FF and FFF peptides leads to different type of nanostructures. Side, top, and cross section views of nanostructures including (a) toroids, (b) nanotubes, (c) spherical nanovesicles, (d) solid nanorods, and (e) solid nanospheres. The CG beads are colored as follows: FF main chain (red), FF side chain (white), FFF main chain (blue), FFF side chain (yellow), water molecule (cyan). (f) Geometry map of the nanostructures formed at seven different FFF mass percentages, plotted using RgY/RgX and RgZ/RgX as two coordinates, where RgX, RgY, and RgZ are the radii of gyration (Rg) of a nanoassembly with respect to the three principle axes X, Y, and Z. RgX, RgY, and RgZ are set in order from smallest to largest Rg values. The dotted line indicates the diagonal of the map. All the data are separated into three different clusters that are highlighted by three green circles.
obtained by using this force field in combination with standard MARTINI water model, are consistent with experimental observations for both FF and FFF assemblies, indicating that the MARTINI 2.1 force field is suitable for the study of both FF and FFF self-assembly. This force-field has also been validated to be an appropriate model for studying the self-assembly of short peptides38,40−42,48−50 and peptide amphiphiles.51 Our 290 independent microsecond-scale MD simulations demonstrate that FF and FFF peptides can not only coassemble into ordered nanostructures with classic or canonical geometries, including hollow nanotubes and spherical nanovesicles, and solid nanorods and nanospheres, but also co-organize into toroids with noncanonical topology. The toroid nanostructure, which is doughnut-shaped and filled with water molecules, is a new type of nanostructure. It is often observed at a FFF mass percentage ranging from 0.167 to 0.33. The simulationpredicted formation of nanotoroids is further experimentally confirmed by scanning electron microscopy. We find that FFF peptides facilitate the formation of negative-curvature surfaces of nanostructures and the presence of FFF peptides weakens the overall peptide-water interactions, thus modulating the formation of different morphologies. Strikingly, the FFFinduced hollow-solid FF-FFF nanostructural transformation produced a sigmoidal curve, indicative of a discontinuous transition behavior. The role of peptide−peptide and peptide− water interactions on hollow-solid structural transition and toroid formation is discussed.
independent MD runs and the representative snapshot of the most-populated nanostructure. The snapshots in the last two columns demonstrate that a rich variety of ordered nanostructures were observed in our simulations, including toroids, nanotubes, spherical nanovesicles, ellipsoidal nanovesicles (ellipsoids), nonorods, and nanospheres. Distinct from the self-assembly of FF or FFF alone, which generates almost solely classic nanovesicles and nanotubes for FF peptides38 and solid nanospheres and nanorods for FFF peptides,40 the coassembly of FF and FFF leads to more diverse of nanostructures. Particularly, the nanotoroid with its geometry and topology distinctly different from the classic cylindrical and spherical-like nanoassemblies, which was rarely observed in the self-assembly of FF or FFF peptides, is often seen in the coassembly of FF and FFF peptides. Table 1 shows that nanotoroids are the most-abundant nanoassemblies at PFFF = 0.167−0.33. Hollow structures are dominant when PFFF < 0.5, while solid structures become the most abundant when PFFF ≥ 0.5. Three different views of the five different FF-FFF coassembles are given in Figure 1. The toroid structures are doughnut-shaped and filled with water molecules (Figure 1a). Nanotubes (Figure 1b) and spherical nanovesicles (Figure 1c) are also filled with water, while nonorods (Figure 1d) and nanospheres (Figure 1e) contain much less water molecules and are considered as solid nanostructures. We characterize the shape of the nanostructures by constructing the geometry map using RgY/RgX and RgZ/ RgX as two reaction coordinates, where RgX, RgY, and RgZ are the radii of gyration (Rg) of a given nanoassembly with respect to, respectively, the three principle axes X, Y, and Z of the nanostructure. They are defined as follows:
RESULTS AND DISCUSSION FF and FFF Peptides Coassemble into Varied Nanostructures in a Ratio-Dependent Manner. We investigated the coassembly of FF and FFF peptides for 16 different peptide systems with the FFF mass percentage, PFFF, ranging from 0 to 1, starting from a homogeneous mixture of FF and FFF peptides in water for each system using extensive CG-MD simulations. A total of 290 independent 1.2−4.8 μs MD runs show the spontaneous formation of cylindrical or spherical nanostructures in morphology. These nanostructures can be classified as hollow and solid assemblies according to the number of water molecules buried inside the nanostructure. We give in Table 1, for each peptide system, the number of MD runs that lead to hollow or solid nanostructures in 10−40
RgX =
RgZ =
∑i mi(yi 2 + zi 2) ∑i mi
, RgY =
∑i mi(xi 2 + zi 2) ∑i mi
,
∑i mi(xi 2 + yi 2 ) ∑i mi
where mi and (xi, yi, zi) are, respectively, the mass and the coordinates of peptide bead i and the summations are over all beads in a given nanoassembly. RgX, RgY, and RgZ are set in 8318
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
Figure 2. Analysis of toroidal nanovesicle formation pathway. (a) Representative pathway from one of CG MD trajectories at PFFF = 0.2. The cross-section views of the nanostructures at t = 232, 401, and 2970 ns are also given. The CG beads are colored as follows: FF main chain (red), FF side chain (white), FFF main chain (blue), and FFF side chain (yellow). For clarity, water beads are not shown. The time evolution of (b) Rg and (c) SASA of peptide system reflects a quick assembly process. RgX, RgY, and RgZ correspond to the three Rg values arranged in order from smallest to largest with respect to the principle axes X, Y, and Z of the nanoassembly.
of bilayered assemblies (including small clusters, a large sheetlike structure, and a nanovesicle) within 50 ns. After that, the large bilayered sheet bends in three dimensions and adopts a hemispherical shape at t = 100 ns. The hemispherical assembly evolves with time and changes into a spherical nanovesicle at t = 167 ns. Another two spherical nanovesicles are also observed at the same time albeit with different sizes. Thereafter, two of the three vesicles start to fuse and merge into one assembly at t = 175 ns. This is followed by the joining of the third vesicle (see the snapshot at t = 194 ns). A large assembly is formed at t = 232 ns, with a tiny hole in its center. The cross section view of the FF-FFF coassembly shows that it is a toroid-like structure with an imperfect surface. After hundreds of nanoseconds self-organization, a well-organized toroid is formed at t = 2970 ns, with the hole becoming more apparent. The formation process of FF-FFF toroidal nanovesicles can further be monitored by the time evolutions of Rg and solvent accessible surface area (SASA) of the peptide system. As shown in Figure 2b, the three Rg values undergo a sharp decrease in the first 100 ns, reflecting a fast initial aggregation process. The RgY and RgX values gradually decrease and both converge to a value of ∼3.0 nm at t = 1200 ns, but smaller than the RgZ value of 3.9 nm. The similar value of RgY and RgX indicates a circular shape of the toroid in its side view. Similarly, the SASA quickly reduces to 600 nm2 from its initial value of 2600 nm2 within the first 500 ns (see Figure 2c). It fluctuates around this value in the remaining 2500 ns of the simulation, accompanied by the self-organization of the nanotoroid. The FF-FFF coassembly pathways of other nanostructures (including nanovesicles, nanotubes, nanospheres and nanorods) are given in Figures S2−S5, which show that these nanoassemblies follow distinct coassembly processes. To verify the existence of this new toroid nanostructure, we carried out a coassembly experiment of FF and FFF, and examined the products using scanning electron microscopy
order from smallest to largest Rg values. We then projected all nanostructures obtained at seven different PFFF in two reaction coordinates: RgY/RgX and RgZ/RgX. Figure 1f shows that all the data points are well separated into three different clusters: 1, 2, and 3, as green-circled in Figure 1f. Cluster 1 corresponds to nanotubes or nanorods whose RgZ and RgY are almost the same but much larger than RgX. Cluster 2 corresponds to ellipsoids whose RgZ and RgY are almost the same but larger than RgX. The data on the diagonal line in cluster 3 correspond to sphere-like nanovesicles or nanospheres, while those in the upper left of the diagonal correspond to toroids or ellipsoids. The geometry map reflects the shape diversity of the nanostructures. We also examined the side chain orientations of FF and FFF peptides in the five different nanoassemblies (toroids, nanotubes, spherical nanovesicles, nanorods, and nanospheres) formed at PFFF = 0.2 by calculating the distribution of the dihedral angle α and β (see the definition in Figure S1 insets). The probability distribution of dihedral angle in Figure S1 reveals that FF mostly adopts a rigid planar FF conformation, while FFF tends to have a flexible conformation, which can readily self-adjust to form different nanostructures. Computational and Experimental Observation of Nanotoroids. Among all the nanostructures obtained here, the nanotoroid structure is particularly interesting as it has a distinct geometry and topology from the classic cylindrical or spherical nanostructures formed by FF or FFF alone.38,40 Moreover, the peptide toroidal nanovesicles have rarely been detected by experiments or by MD simulations of FF26,38,41 or FFF peptide.34,40,42 To probe the coassembly pathway of FF and FFF into toroids at molecular level, we analyzed one of the toroid formation trajectories for PFFF = 0.2 (see Figure 2). It can be seen from Figure 2a that starting from the initially randomly distributed 576 FF and 96 FFF peptide chains in a water box, the peptides aggregate rapidly and form various sizes 8319
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
right, indicating that FFF peptides have a preference to be located in the inner leaflet of the nanostructures. This can be seen clearly from the cross section view of the nanostructure in the inset of Figure 4. The inner surface is concave with a negative curvature, while the outer surface is convex with a positive curvature. Thus, peptides in the inner leaflet are packed more condensely than those in the outer leaflet, which can reduce the water exposure of peptides. Compared with FF peptides, FFF peptides preferentially reside in the negativecurvature leaflet due to their relatively stronger hydrophobicity. Thus, it is energetically favorable to bury their side chains as much as possible while minimizing the total SASA of the nanostructure. A spherical vesicle has the smallest outer surface area but has a larger inner cavity. Compared with a spherical vesicle, a nanotube is filled with fewer water molecules but has a larger outer surface area. Compromisingly, a toroid, whose outer surface area and inner cavity are between those of a spherical vesicle and nanotube, becomes the most favorable nanostructure at an appropriate FFF mass percentage, i.e., 0.167−0.33. It is conceivable that with more FFF molecules added in the solution, the toroid structure would disappear and transit into solid nanospheres/nanorods that have a much smaller SASA. The Mechanisms of Hollow-Solid Nanostructure Transition and Nanotoroid Formation. Understanding the nature of FFF-induced hollow-solid nanostructure transition and the underlying driving force is crucial for achieving controllable FF-FFF coassembly. To this aim, we first plotted in Figure 5a the percentage of hollow or solid nanostructures as a function of FFF mass percentage (PFFF). Figure 5a shows that, with the increase of PFFF, the percentage of hollow structures decreases, while that of solid nanostructure increases. Hence the content of FFF delicately modulates the coassembly of FF and FFF peptide mixtures, making the formation of FF-FFF nanostructures controllable. Interestingly, the variation of hollow or solid nanostructure percentage with PFFF (in the range of 0−1.0) displays a sigmoidal curve, revealing a discontinuous hollow-solid structure transition. Peptide−solvent interaction usually plays an important role in peptide self-assembly. FF and FFF are short peptides with hydrophilic main chains and hydrophobic side chains, making them somewhat amphiphilic molecules. Thus, it is interesting to explore the role played by peptide−water interactions in the FFF-induced hollow-solid nanostructure transition. We calculated the number of water beads within 0.65 nm from peptide beads as a function of FFF mass percentage. With the increase of PFFF, the number of water molecules that have molecular
(SEM). First, FF and FFF were dissolved separately to a concentration of 100 mg/mL in 1,1,1,3,3,3-hexafluoro-2propanol (HFIP) and then mixed together to the desired PFFF. The assembly process was induced by introducing the dissolved peptide solution into water to a total concentration of 4 mg/mL in 4% HFIP. A series of PFFF ranging from 0.05 to 0.5 was prepared and imaged. As shown in Figure 3, the SEM
Figure 3. SEM of FF-FFF coassembly formed at PFFF of (a) 0.33 and (b) 0.5. Peptides were dissolved in HFIP and then mixed to a desired ratio. Assembly was triggered using solvent-switch into an aqueous solution.
photo clearly shows that the formation of toroid nanostructures with varying inner radii. Toroidal structures were observed at a PFFF value of 0.2, 0.33, and even 0.5. We note the toroids observed experimentally are mostly much larger than those generated in our MD simulations. This discrepancy is probably due to the difference in the sizes of systems studied in our simulations and experiments. Nevertheless, they exhibit the same shape and topology. FFF Peptides Have a Preference to Be Located in the Inner Leaflet of Hollow Nanostructures. It is essential to reveal the feature of spatial distribution of FFF molecules in the hollow FF-FFF coassemblies, in particular, in toroid nanostructures. To this aim, we calculated the radial distribution function (RDF) of main chain beads and side chain beads (using the mass center of each aromatic ring) of FFF molecules in the hollow nanostructures including toroids, spherical-like nanovesicles, and nanotubes for PFFF = 0.25. The reason that we chose this PFFF value is that toroids are mostly populated around a FFF mass percentage of 0.25 (see Table 1). As shown in Figure 4, the RDFs of FFF main chain beads display two peaks for all of the three nanostructures, while those of side chain beads have only one peak located between the two main chain peaks. The two main chain peaks, from left to right, correspond, respectively, to the locations of the inner and outer leaflet of the bilayered wall of hollow nanostructures. The peak in the left of the RDF curve is much higher than that in the
Figure 4. RDF of the main chain and side chain beads of FFF molecules in hollow FF-FFF nanostructures including toroid, spherical-like nanovesicles, and nanotubes. As depicted by the axis arrow in the inset, the RDF is calculated with respect to the central circular axis of the toroid tube (a) and the principal axis of a spherical nanovesicle (b) and a nanotube (c). The snapshots in the insets are used only to show how RDF was calculated, and they were generated by cutting out a thin slice of the nanostructure. 8320
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
Figure 5. Analysis of mechanism of FFF-induced hollow-solid structure transition and toroid formation. We plot, as a function of FFF mass percentage, (a) percentage of hollow or solid structure; (b) number of water beads within 0.65 nm of peptide (including FF and FFF) atoms; (c) ratio of hollow to solid nanostructures (for clarity, data with solely hollow or solid structures were excluded). We fitted the plot in panel (c) using an exponential function y = a*exp(−bx) + c (a = 222.16, b = 15.01, c = 0.41) and the R2 value is 0.9659; (d) ratio of interacting FF to FFF molecules (i.e., the ratio of the number of FF residues to the number of available FFF residues. For example, 10 FF molecules that have atomic contact with eight FFF molecules gives a FF:FFF ratio of (10*2)/(8*3) = 0.83). We fitted the plot in panel (d) using an exponential function y = a*exp(−bx) + c (a = 4.09, b = 3.18, c = 0.09) and the R2 value is 0.9998; (e) percentage of the toroid nanostructure and (f) number of water beads within 0.65 nm of FF (red) or FFF (green) molecules, normalized by the total residue number of 1440. The last 150 ns data of each trajectory were used for analysis.
We also plot the percentage of toroids as a function of PFFF. It can be seen from Figure 5e that the percentage of toroid structures has a maximum value of ∼60% at PFFF = 0.25. The FFF mass percentage that goes much beyond or under 0.25 impedes the toroid nanostructure formation. Figure 5a shows that the percentage of hollow structures decreases with the increase of PFFF and reduces to 85% at PFFF = 0.25. These data indicate at PFFF = 0.25, most of the hollow nanostructures are toroids. Thus, the toroid appears to be an obligatory intermediate for the FFF-induced hollow-solid nanostructure transformation. As discussed above, the number of water molecules that have molecular contacts with FF-FFF assemblies (i.e., peptide−water interaction) decreases as PFFF increases (Figure 5b). It is of great interest to explore how the hydration capability of each type of peptide (i.e., FF−water or FFF−water interaction) changes with the increase of PFFF. To this end, we calculated the number of water beads within 0.65 nm of FF or FFF molecules. The number of water molecules per FFF or FF residue as a function of PFFF is given in Figure 5f. With the increase of PFFF, the number of water molecules around FF increases, while that around FFF decreases (Figure 5f). The two curves in Figure 5f intersect each other at PFFF = 0.25, which indicates that FFF and FF have similar hydration capability at this PFFF. Surprisingly, the percentage of toroids reaches its maximum value around PFFF = 0.25 (Figure 5e). This implies that it is most favorable for the peptides to form toroid nanostructures when the extent of competition between FF−water and FFF− water interactions becomes similar. Thus, we propose that the extent of competition between FF−water and FFF−water interactions determines the formation of toroid structures
contacts with FF-FFF nanostructures monotonically decreases (Figure 5b), indicating that increasing FFF mass percentage weakens peptide−water interactions. FFF peptide has a larger bulky hydrophobic side chain than FF peptide, and it is thermodynamically favorable to bury its side chains from being exposed to water. Therefore, the introduction of more FFF molecules into the FF-FFF mixture facilitates the formation of assemblies that have a reduced water accessible surface area, which shifts the conformations toward solid nanostructures (see Figure 5a). Interestingly, dividing the percentage of hollow structure (i.e., the red curve in Figure 5a) by that of a solid one (i.e., the green curve in Figure 5a) generates an exponential decay curve (Figure 5c). Unexpectedly, the ratio of FF:FFF molecules (that contribute to FF-FFF molecular contacts) with the increase of PFFF also displays a perfectly smooth exponential decay (Figure 5d). This exponential decay reveals that FFF or FF molecules have an increased tendency to interact with themselves rather than with FF or FFF molecules in the FF-FFF mixture as PFFF increases. That is to say FFF peptides preferentially cluster together and are nonuniformly distributed in the FF-FFF coassemblies. This can be clearly seen from the snapshot in Figure 2. By comparing Figure 5c with 5b and d, we find that the change trend of the hollow:solid ratio with the increase of PFFF (Figure 5c) is quite similar to that of interacting FF:FFF ratio (Figure 5d), but different from that of the number of water molecules interacting with the coassemblies (Figure 5b). The shape similarity between the two curves in Figure 5c and d indicates that the hollow-solid nanostructure transition is mostly dominated by FF−FFF interactions, although peptide−water interactions also play roles. 8321
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
examine the effect of Martini force field on the distribution of the dihedral angles of FF molecules, we calculated the distribution of the side chain−backbone−backbone−side chain dihedral angle (β1−α1−α2−β2) of FF molecules in the all-atom FF vesicles (Figure S6a) obtained in our previous work38 and compared it with the dihedral distribution of FF peptides in CG vesicles (Figure S6b). Figure S6 shows that the dihedral angle distribution of FF molecules in the all-atom and CG models is quite similar overall, with insignificant differences in the range of 90−180°. These results suggest that for the ultrashort di- and triphenylalanine peptides, the Martini force field does not introduce significant errors on the dihedral angle distribution. We also performed MD simulations using MARTINI 2.2 force field55 for the system with a PFFF = 0.25 and PFFF = 0.0. The simulation results for PFFF = 0.25 are given in Figure S7. However, ordered nanostructures were not observed within 1200 ns. The reason for using CG-MD simulations and MARTINI 2.1 force field is given in more detail in the Supporting Information. Initiated from a homogeneous mixture of FF-FFF peptides in water, 10−40 independent MD runs were performed for each system. The simulation time of each MD run is 1.2−4.8 μs long. Each peptide system consists of 1440 phenylalanine (F) amino acid residues. The mass percentage of FFF peptide (PFFF) and the number of FF or FFF peptide molecules are listed in Table 1. The simulation box size is 17 × 17 × 17 nm3, and the peptide concentration is ∼74 mg/mL. The number of water beads is ∼38915 for all the systems. All simulations were performed in the NPT ensemble using the GROMACS 4.5.3 software.56 More details about the MARTINI force field and CG-MD simulations are given in Supporting Information. Electron Microscopy Imaging. The two peptides were dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) at a concentration of 100 mg/mL. Then, both solutions were mixed together at different weight ratios and were introduced into water, in order to trigger the assembly process. Samples with PFFF of 0.5, 0.33, 0.2, and 0.05 were then coated with Cr and viewed using a JSM-6700 field-emission HR-SEM (Jeol, Tokyo, Japan), equipped with a cold field emission gun, operating at 10 kV. The final peptide concentration in all samples was 4 mg/mL in 4% HFIP solution.
Here we demonstrate that the coassembly of FF and FFF--two peptides consisting of naturally occurring amino acids yields diverse architectures including both canonical and noncanonical nanoassemblies. Coassembling of peptides with artificially synthesized peptide analogues has also been reported. For example, Yuran et al. showed that FF and its tert-butyl dicarbonate (Boc)-modified analogue (i.e., Boc-FF, which forms nanospheres) coassembled into a construction of beaded strings.52 Maity et al. investigated the coassembly of FF and Fmoc-L-DOPA(acetonated)-D-Phe-OMe and found that the two peptides formed spherical assemblies similar in morphology to either red or white blood cells depending on the peptide concentrations.53 Taken together, these results demonstrate that the approach of coassembly would serves as a potentially effective strategy to increase the structural diversity of peptide nanomaterials by taking advantage of the diversity of existing assembly building blocks. The methods used in this work for characterizing the geometry map and peptide−peptide and peptide−water interactions can be also extended to many other short peptides, such as IF, FFF, WW, and WWW.
CONCLUSIONS By conducting extensive MD simulations on a microsecond time scale, we have studied the coassembly of FF and FFF peptides at various mass ratios. Our MD simulations show that the coassembly of FF and FFF can lead to various nanostructures in a ratio-dependent manner. The toroid nanostructures, whose geometry and topology are distinct from those formed by self-assembly of FF or FFF solely, were often observed in our MD simulations for PFFF in the range of 0.167−0.33. Our toroid formation predictions were confirmed by SEM imaging of FF-FFF coassemblies at different ratios. Analysis of the spatial distribution of FFF molecules in hollow nanostructures demonstrates that FFF favors negative curvature where its bulky hydrophobic side chains can be more closely packed and protected from exposure to water. In line with this nature, the increase of FFF mass percentage in the FFF-FF mixture monotonically decreases the peptide−water interactions. As a result, the formation probability of different nanostructures can be controlled by modulating the FFF:FF mass ratio. The transformation from hollow to solid FF-FFF coassemblies displays a discontinuous transition feature. The toroid structures play a role as an intermediate structure between hollow and solid nanostructure transition. Thorough investigation of the underlying driving force of the coassembly reveals that the hollow-solid structural transition is mostly dominated by the FF−FFF interaction, while the formation of toroid nanostructure is determined by the extent of competition between FF−water and FFF−water interactions. Thus, the FFF:FF mass ratio can be a controllable coassembly regulator. The molecular mechanism revealed here may have a significant contribution to understanding the process of peptide coassembly into nanostructures and may provide a molecular basis for rational design of biomaterials.
ASSOCIATED CONTENT S Supporting Information *
The SI contains The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/ acsnano.6b02739. Figures S1−S7, details of MARTINI force field and CGMD simulations, and the reason for using CG-MD simulations and MARTINI 2.1 force field (PDF)
AUTHOR INFORMATION Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. *E-mail:
[email protected]. Author Contributions
C.G. and R.Q. performed the simulations; Z.A. and L.A. performed the experiments; all authors analyzed the data; G.W., Q.Z., C.G., and R.Q. conceived the research; R.Q., C.G., Z.A., Q.Z., G.W., and E.G. cowrote the paper. Notes
MATERIALS AND METHODS
The authors declare no competing financial interest.
MD Simulations. A total of 16 FF-FFF peptide systems have been simulated by CG-MD simulations with the MARTINI 2.1 force field46,47 in combination with a standard water model, with different FFF mass percentage for each system. Very recently, it was reported that significant errors could be introduced by the Martini force field on the distribution of the side chain−backbone−backbone−side chain dihedral angles, especially for β-sheet peptides and proteins.54 To
ACKNOWLEDGMENTS We thank B. Ma for insightful discussion. G.W. acknowledges the financial support from the MOST of China (grant no.: 2016YFA0501702) and the NSF of China (grant nos.: 91227102, 11274075, and 11674065). E.G. acknowledges the 8322
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano
(20) Han, T. H.; Lee, W. J.; Lee, D. H.; Kim, J. E.; Choi, E. Y.; Kim, S. O. Peptide/Graphene Hybrid Assembly into Core/shell Nanowires. Adv. Mater. 2010, 22, 2060−2064. (21) Lee, J. S.; Yoon, I.; Kim, J.; Ihee, H.; Kim, B.; Park, C. B. SelfAssembly of Semiconducting Photoluminescent Peptide Nanowires in the Vapor Phase. Angew. Chem., Int. Ed. 2011, 50, 1164−1167. (22) Han, T. H.; Oh, J. K.; Park, J. S.; Kwon, S.-H.; Kim, S.-W.; Kim, S. O. Highly Entangled Hollow TiO2 Nanoribbons Templating Diphenylalanine Assembly. J. Mater. Chem. 2009, 19, 3512−3512. (23) Huang, R.; Qi, W.; Su, R.; Zhao, J.; He, Z. Solvent and Surface Controlled Self-assembly of Diphenylalanine Peptide: from Microtubes to Nanofibers. Soft Matter 2011, 7, 6418−6418. (24) Li, Q.; Jia, Y.; Dai, L.; Yang, Y.; Li, J. Controlled Rod Nanostructured Assembly of Diphenylalanine and Their Optical Waveguide Properties. ACS Nano 2015, 9, 2689−2695. (25) Smith, A. M.; Williams, R. J.; Tang, C.; Coppo, P.; Collins, R. F.; Turner, M. L.; Saiani, A.; Ulijn, R. V. Fmoc-Diphenylalanine Self Assembles to a Hydrogel via a Novel Architecture Based on π−π Interlocked β-Sheets. Adv. Mater. 2008, 20, 37−41. (26) Yan, X.; Zhu, P.; Li, J. Self-assembly and Application of Diphenylalanine-based Nanostructures. Chem. Soc. Rev. 2010, 39, 1877−1890. (27) Zhu, P. L.; Yan, X. H.; Su, Y.; Yang, Y.; Li, J. B. Solvent-induced Structural Transition of Self-assembled Dipeptide: from Organogels to Microcrystals. Chem. - Eur. J. 2010, 16, 3176−3183. (28) Marchesan, S.; Vargiu, A.; Styan, K. The Phe-Phe Motif for Peptide Self-Assembly in Nanomedicine. Molecules 2015, 20, 19775− 19788. (29) Sedman, V. L.; Adler-Abramovich, L.; Allen, S.; Gazit, E.; Tendler, S. J. B. Direct Observation of the Release of Phenylalanine from Diphenylalanine Nanotubes. J. Am. Chem. Soc. 2006, 128, 6903− 6908. (30) Hauser, C. A. E.; Zhang, S. Nanotechnology: Peptides as Biological Semiconductors. Nature 2010, 468, 516−517. (31) Yan, X.; He, Q.; Wang, K.; Duan, L.; Cui, Y.; Li, J. Transition of Cationic Dipeptide Nanotubes into Vesicles and Oligonucleotide Delivery. Angew. Chem., Int. Ed. 2007, 46, 2431−2434. (32) Adler-Abramovich, L.; Vaks, L.; Carny, O.; Trudler, D.; Magno, A.; Caflisch, A.; Frenkel, D.; Gazit, E. Phenylalanine Assembly into Toxic Fibrils Suggests Amyloid Etiology in Phenylketonuria. Nat. Chem. Biol. 2012, 8, 701−706. (33) Do, T. D.; Kincannon, W. M.; Bowers, M. T. Phenylalanine Oligomers and Fibrils: The Mechanism of Assembly and the Importance of Tetramers and Counterions. J. Am. Chem. Soc. 2015, 137, 10080−10083. (34) Tamamis, P.; Adler-Abramovich, L.; Reches, M.; Marshall, K.; Sikorski, P.; Serpell, L.; Gazit, E.; Archontis, G. Self-Assembly of Phenylalanine Oligopeptides: Insights from Experiments and Simulations. Biophys. J. 2009, 96, 5020−5029. (35) Han, T. H.; Ok, T.; Kim, J.; Shin, D. O.; Ihee, H.; Lee, H. S.; Kim, S. O. Bionanosphere Lithography via Hierarchical Peptide Selfassembly of Aromatic Triphenylalanine. Small 2010, 6, 945−951. (36) German, H. W.; Uyaver, S.; Hansmann, U. H. E. Self-assembly of Phenylalanine-based Molecules. J. Phys. Chem. A 2015, 119, 1609− 1615. (37) Yan, X. H.; Cui, Y.; He, Q.; Wang, K. W.; Li, J. B.; Mu, W. H.; Wang, B. L.; Ou-yang, Z. C. Reversible Transitions between Peptide Nanotubes and Vesicle-like Structures Including Theoretical Modeling Studies. Chem. - Eur. J. 2008, 14, 5974−5980. (38) Guo, C.; Luo, Y.; Zhou, R. H.; Wei, G. H. Probing the Selfassembly Mechanism of Diphenylalanine-based Peptide Nanovesicles and Nanotubes. ACS Nano 2012, 6, 3907−3918. (39) Jeon, J.; Mills, C. E.; Shell, M. S. Molecular Insights into Diphenylalanine Nanotube Assembly: All-atom Simulations of Oligomerization. J. Phys. Chem. B 2013, 117, 3935−43. (40) Guo, C.; Luo, Y.; Zhou, R.; Wei, G. Triphenylalanine Peptides Self-assemble into Nanospheres and Nanorods that Are Different from the Nanovesicles and Nanotubes Formed by Diphenylalanine Peptides. Nanoscale 2014, 6, 2800−2811.
support by grants from the Israeli National Nanotechnology Initiative and Helmsley Charitable Trust for a focal technology area on nanomedicines for Personalized Theranostics. Simulations were performed at the National Supercomputing Center in Guangzhou and the National High Performance Computing Center of Fudan University.
REFERENCES (1) Zhang, S. Fabrication of Novel Biomaterials Through Molecular Self-assembly. Nat. Biotechnol. 2003, 21, 1171−1178. (2) 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, 594−601. (3) Cui, H.; Webber, M. J.; Stupp, S. I. Self-assembly of Peptide Amphiphiles: From Molecules to Nanostructures to Biomaterials. Biopolymers 2010, 94, 1−18. (4) Ikezoe, Y.; Washino, G.; Uemura, T.; Kitagawa, S.; Matsui, H. Autonomous Motors of a Metal−organic Framework Powered by Reorganization of Self-assembled Peptides at Interfaces. Nat. Mater. 2012, 11, 1081−1085. (5) Tsai, C. J.; Zheng, J.; Nussinov, R. Designing a Nanotube using Naturally Occurring Protein Building Blocks. PLoS Comput. Biol. 2006, 2, e42. (6) Tsai, C.-J.; Zheng, J.; Zanuy, D.; Haspel, N.; Wolfson, H.; Alemán, C.; Nussinov, R. Principles of Nanostructure Design with Protein Building Blocks. Proteins: Struct., Funct., Genet. 2007, 68, 1−12. (7) Makin, O. S.; Sikorski, P.; Serpell, L. C. Diffraction to Study Protein and Peptide Assemblies. Curr. Opin. Chem. Biol. 2006, 10, 417−422. (8) Ulijn, R. V.; Smith, A. M. Designing Peptide Based Banomaterials. Chem. Soc. Rev. 2008, 37, 664−675. (9) Larini, L.; Shea, J.-E. Role of β-hairpin Formation in Aggregation: the Self-assembly of the Amyloid-β(25−35) Peptide. Biophys. J. 2012, 103, 576−86. (10) Kang, S. G.; Li, H.; Huynh, T.; Zhang, F.; Xia, Z.; Zhang, Y.; Zhou, R. Molecular Mechanism of Surface-assisted Epitaxial Selfassembly of Amyloid-like Peptides. ACS Nano 2012, 6, 9276−82. (11) Miller, Y.; Ma, B.; Nussinov, R. Polymorphism in Self-assembly of Peptide-based Beta-hairpin Contributes to Network Morphology and Hydrogel Mechanical Rigidity. J. Phys. Chem. B 2015, 119, 482− 90. (12) Gorbitz, C. H. Nanotube Formation by Hydrophobic Dipeptides. Chem. - Eur. J. 2001, 7, 5153−5159. (13) Reches, M.; Gazit, E. Casting Metal Nanowires within Discrete Self-assembled Peptide Nanotubes. Science (Washington, DC, U. S.) 2003, 300, 625−627. (14) Reches, M.; Gazit, E. Controlled Patterning of Aligned Selfassembled Peptide Nanotubes. Nat. Nanotechnol. 2006, 1, 195−200. (15) Hill, R. J. A.; Sedman, V. L.; Allen, S.; Williams, P.; Paoli, M.; Adler-Abramovich, L.; Gazit, E.; Eaves, L.; Tendler, S. J. B. Alignment of Aromatic Peptide Tubes in Strong Magnetic Fields. Adv. Mater. 2007, 19, 4474−4479. (16) Song, Y. J.; Challa, S. R.; Medforth, C. J.; Qiu, Y.; Watt, R. K.; Pena, D.; Miller, J. E.; van Swol, F.; Shelnutt, J. A. Synthesis of Peptide-nanotube Platinum-nanoparticle Composites. Chem. Commun. 2004, 1044−1045. (17) Yan, X. H.; He, Q.; Wang, K. W.; Duan, L.; Cui, Y.; Li, J. B. Transition of Cationic Dipeptide Nanotubes into Vesicles and Oligonucleotide Delivery. Angew. Chem., Int. Ed. 2007, 46, 2431−2434. (18) Charalambidis, G.; Kasotakis, E.; Lazarides, T.; Mitraki, A.; Coutsolelos, A. G. Self-Assembly into Spheres of a Hybrid Diphenylalanine−Porphyrin: Increased Fluorescence Lifetime and Conserved Electronic Properties. Chem. - Eur. J. 2011, 17, 7213−7219. (19) Kim, J.; Han, T. H.; Kim, Y. I.; Park, J. S.; Choi, J.; Churchill, D. C.; Kim, S. O.; Ihee, H. Role of Water in Directing Diphenylalanine Assembly into Nanotubes and Nanowires. Adv. Mater. 2010, 22, 583− 587. 8323
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
Article
ACS Nano (41) Frederix, P.; Ulijn, R. V.; Hunt, N. T.; Tuttle, T. Virtual Screening for Dipeptide Aggregation: Toward Predictive Tools for Peptide Self-assembly. J. Phys. Chem. Lett. 2011, 2, 2380−2384. (42) Frederix, P. W. J. M.; Scott, G. G.; Abul-Haija, Y. M.; Kalafatovic, D.; Pappas, C. G.; Javid, N.; Hunt, N. T.; Ulijn, R. V.; Tuttle, T. Exploring the Sequence Space for (Tri-)peptide Selfassembly to Design and Discover New Hydrogels. Nat. Chem. 2015, 7, 30−37. (43) Gazit, E. Self-assembled Peptide Nanostructures: the Design of Molecular Building Blocks and Their Technological Utilization. Chem. Soc. Rev. 2007, 36, 1263−1269. (44) Do, T. D.; Bowers, M. T. Diphenylalanine Self Assembly: Novel Ion Mobility Methods Showing the Essential Role of water. Anal. Chem. 2015, 87, 4245−4252. (45) Rissanou, A. N.; Georgilis, E.; Kasotakis, E.; Mitraki, A.; Harmandaris, V. Effect of Solvent on the Self-assembly of Dialanine and Diphenylalanine Peptides. J. Phys. Chem. B 2013, 117, 3962−3975. (46) Marrink, S. J.; Risselada, H. J.; Yefimov, S.; Tieleman, D. P.; de Vries, A. H. The MARTINI Force Field: Coarse Grained Model for Biomolecular Simulations. J. Phys. Chem. B 2007, 111, 7812−7824. (47) Monticelli, L.; Kandasamy, S. K.; Periole, X.; Larson, R. G.; Tieleman, D. P.; Marrink, S. J. The MARTINI Coarse-grained Force Field: Extension to Proteins. J. Chem. Theory Comput. 2008, 4, 819− 834. (48) Sun, Y.; Qian, Z.; Guo, C.; Wei, G. Amphiphilic Peptides A6K and V6K Display Distinct Oligomeric Structures and Self-Assembly Dynamics: A Combined All-Atom and Coarse-Grained Simulation Study. Biomacromolecules 2015, 16, 2940−2949. (49) Marrink, S. J.; Tieleman, D. P. Perspective on the Martini Model. Chem. Soc. Rev. 2013, 42, 6801−6822. (50) Periole, X.; Marrink, S.-J. The Martini Coarse-grained Force Field. Methods Mol. Biol. 2013, 924, 533−565. (51) Lee, O. S.; Cho, V.; Schatz, G. C. Modeling the Self-assembly of Peptide Amphiphiles into Fibers using Coarse-grained Molecular Dynamics. Nano Lett. 2012, 12, 4907−4913. (52) Yuran, S.; Razvag, Y.; Reches, M. Coassembly of Aromatic Dipeptides into Biomolecular Necklaces. ACS Nano 2012, 6, 9559− 9566. (53) Maity, S.; Nir, S.; Reches, M. Co-assembly of Aromatic Dipeptides into Spherical Structures that Are Similar in Morphology to Red and White Blood Cells. J. Mater. Chem. B 2014, 2, 2583−2591. (54) Herzog, F. A.; Braun, L.; Schoen, I.; Vogel, V. Improved Side Chain Dynamics in MARTINI Simulations of Protein-Lipid Interfaces. J. Chem. Theory Comput. 2016, 12, 2446−2458. (55) de Jong, D. H.; Singh, G.; Bennett, W. F.; Arnarez, C.; Wassenaar, T. A.; Schäfer, L. V.; Periole, X.; Tieleman, D. P.; Marrink, S. J. Improved Parameters for the Martini Coarse-Grained Protein Force Field. J. Chem. Theory Comput. 2013, 9, 687−697. (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.; et al. GROMACS 4.5: A High-Throughput and Highly Parallel Open Source Molecular Simulation Toolkit. Bioinformatics 2013, 29, 845−854.
8324
DOI: 10.1021/acsnano.6b02739 ACS Nano 2016, 10, 8316−8324
