SELF-ASSEMBLING TRIPEPTIDES

Abstract

The present invention relates to a method of predicting the propensity of tripeptides to from aggregates in solution. The present invention also provides tripeptides which are able to form aggregates in solution, as well as uses thereof. The present invention also provides nanostructures formed by self-aggregation of tripeptides of the present invention. The present invention also provides pH responsive aggregates as well as methods of screening for the ability of a tripeptide to form a pH dependent aggregate or gel.

Claims

1. A solution comprising a self-aggregated tripeptide, the tripeptide having the formula:
A.sub.1-A.sub.2-A.sub.3 wherein A.sub.1 is a hydrogen bond donating amino acid, such as K, R, S, T, or P, H, or W, or F; A.sub.2 is an aromatic amino acid, such as F, Y, W, or H; A.sub.3 is an aromatic amino acid, such as F, Y, or W, or a negatively charged amino acid such as D or E.

2. The solution according to claim 1 wherein, the tripeptide has the formula;
A.sub.1-A.sub.2-A.sub.3 wherein A.sub.1 is K, R, S, T or P; A.sub.2 is F, Y, or W; A.sub.3 is F, Y, W, D or E.

3. A solution comprising a self-aggregated tripeptide, wherein the tripeptide is identified in Table 2.

4. The solution according to any of claims 1-3 wherein the peptide displays a hydrophilicity-adjusted measure of propensity for aggregation (AP.sub.H) value of >0.10, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, or 0.18.

5. The solution according to claim 4 wherein the tripeptide displays a propensity of aggregation (AP) value of <2.0.

6. The solution according to any of claims 1-5 selected from KYF, KFF, KYW KYY and FFD, or KYF, KFF, KYW and KYY.

7. The solution according to any claims 1-5 wherein the self-aggregated excludes the peptides CFF, FFF, VFF, FFV, LFF, VYV, KFG or DFN.

8. A self-aggregate structure or gel obtainable from a solution according to any of claims 1-7.

9. A method of producing a peptide capable of self-aggregation in solution, the method comprising: identifying a peptide by determining a hydrophilicity-adjusted measure of propensity for aggregation (AP.sub.H) for the peptide, the AP.sub.H being determined by adjusting a measure of propensity of aggregation (AP) for the peptide in dependence on a measure of hydrophilicity for the peptide; and synthesising the peptide.

10. A method according to claim 9, wherein the peptide is one of a plurality of peptides, and the identifying of the peptide comprises determining an AP.sub.H for each of the plurality of peptides and identifying at least one of the plurality of peptides in dependence on the determined AP.sub.H for the plurality of peptides.

11. A method of producing a nanostructure comprising a peptide capable of self-aggregation in solution, the method comprising: identifying a peptide by determining a hydrophilicity-adjusted measure of propensity for aggregation (AP.sub.H) for the peptide, the AP.sub.H being determined by adjusting a measure of propensity of aggregation (AP) for the peptide in dependence on a measure of hydrophilicity for the peptide; synthesising the peptide; and allowing the peptide to aggregate in solution.

12. A method according to claim 11, wherein the peptide is one of a plurality of peptides, and the identifying of the peptide comprises determining an AP.sub.H for each of the plurality of peptides and identifying at least one of the plurality of peptides in dependence on the determined AP.sub.H for the plurality of peptides.

13. A method of identifying a peptide for aggregation in solution, the method comprising identifying a peptide by determining a hydrophilicity-adjusted measure of propensity for aggregation (AP.sub.H) for the peptide, the AP.sub.H being determined by adjusting a measure of propensity of aggregation (AP) for the peptide in dependence on a measure of hydrophilicity for the peptide.

14. A method according to any one of claims 9-13, wherein the peptide comprises a tripeptide.

15. A method according to any one of claims 9-14, wherein the identifying of the peptide comprises determining whether the determined AP.sub.H for the peptide meets or exceeds a threshold value for AP.sub.H.

16. A method according to any one of claims 9-15, wherein the peptide is one of a plurality of peptides, and the identifying of the peptide comprises determining an AP.sub.H for each of the plurality of peptides and identifying at least one of the plurality of peptides in dependence on the determined AP.sub.H for the plurality of peptides.

17. A method according to claim 16, wherein the plurality of peptides comprises a plurality of tripeptides.

18. A method according to claim 17, wherein the plurality of peptides comprises substantially all possible tripeptides.

19. A method according to any of claims 16-18, wherein identifying at least one of the plurality of peptides in dependence on the determined AP.sub.H for the plurality of peptides comprises identifying a subset of the plurality of peptides, the subset comprising the peptides having the highest AP.sub.H.

20. A method according to claim 19, wherein the subset of the plurality of peptides comprises the 20, 50, 75 or 100 peptides having the highest AP.sub.H, optionally the 200 peptides having the highest AP.sub.H, further optionally the 400 peptides having the highest AP.sub.H.

21. A method according to claim 19, wherein the subset of the plurality of peptides comprises the 10% of the plurality of peptides having the highest AP.sub.H, optionally the 5% of the plurality of peptides having the highest AP.sub.H, further optionally the 2% of the plurality of peptides having the highest AP.sub.H.

22. A method according to claim 19 wherein identifying the subset of the plurality of peptides comprises identifying all peptides in the plurality of peptides having art AP.sub.H greater than a threshold value for AP.sub.H, or greater than or equal to a threshold value for AP.sub.H.

23. A method according to any one of claims 9-22, further comprising obtaining the AP for the peptide from simulation.

24. A method according to claim 23, wherein obtaining the AP for the peptide from simulation comprises performing a molecular dynamics simulation for the peptide.

25. The method according to claim 24, wherein the AP for the peptide comprises a ratio between a solvent accessible surface area at the beginning of the molecular dynamics simulation and a solvent accessible surface area at the end of the molecular dynamics simulation.

26. The method according to any one of claims 9-25, wherein the measure of hydrophilicity comprises a sum of Wimley-White whole-residue hydrophobicities for amino acids in the peptide.

27. The method of any one of claims 9-26, wherein adjusting the AP for the peptide in dependence on a measure of hydrophilicity for that peptide comprises raising the AP to a power and multiplying by the measure of hydrophilicity.

28. The method of any one of claims 9-27, wherein at least one of the AP and the measure of hydrophilicity is normalised prior to the adjusting of the AP in dependence on the measure of hydrophilicity.

29. The method of any one of claims 9-28, wherein determining AP.sub.H for the peptide comprises using the equation:
AP.sub.H=(AP′).sup.α(logP)′ wherein a is a numerical constant, log P is the measure of hydrophilicity for the peptide, and an apostrophe denotes normalisation.

30. The method of any one of claims 9-29 wherein a has a value between 0.5 and 5, optionally between 1 and 4, further optionally between 1 and 3.

31. A self-aggregating peptide capable of self-aggregation in solution obtainable by the method according to any of claims 13-30.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0080] The present invention will now be further described by way of example and with reference to the figures which show:

[0081] FIG. 1(a) is a schematic diagram of a computing apparatus in accordance with an embodiment; FIG. 1(b) is a flow diagram illustrating a workflow in accordance with an embodiment.

[0082] FIG. 2: shows screening for self-assembling tripeptides. (a) Representation of all 20 gene-encoded amino acids in the MARTINI force field. Different colours represent different types of beads, as indicated by the legend. Image adapted with permission from ref. 33. Copyright 2008 American Chemical Society. (b) 50 ns MD simulation results of GGG, PFF, KFD and KYF tripeptides, showing various levels of aggregation. (c) Aggregation Propensity as a function of hydrophobicity for all 8000 tripeptides. Red diamonds represent all tripeptides with AP>2. Green diamonds represent the top 400 tripeptides from the APH score (AP.sub.H=AP.sup.α.Math.logP, α=2) with the overlapping candidates shown in orange. The arrows point to the data points for GGG, PFF, KFD and KYF.

[0083] FIG. 2 (Cont): shows AP as a function of hydrophilicity for all 8000 tripeptides. Green dots represent the top 400 tripeptides from the hydrophobicity-corrected APH score with alternative values for coefficient α. (d) α=1, (e) α=3.

[0084] FIG. 3: shows from Screening to Design Rules. (a) Normalized AP.sub.H score for all 8000 combinations of three amino acids after a 50 ns simulation. Within every rectangle, the third amino acid is represented by the position of the coloured square at the locations indicated in the legend on the right. (b) Expansion of the highlighted areas in (a) with four peptide entries indicated. (c) Average APH scores of tripeptides with the specific amino acid on the x-axis in the N-terminal (blue), middle (red) and C-terminal position (green). A higher score indicates a higher propensity to aggregate. Amino acids are grouped by aromatic, hydrophilic, cationic and anionic side chains. (d). AP score for all 8000 combinations of three amino acids after a 50 ns simulation. Within every rectangle, the third amino acid is represented by the position of the coloured square at the locations indicated in the legend on the right.

[0085] FIG. 4 shows the Characterization of Selected Tripeptides. (a-c) TEM images of KYF, KFD and PFF tripeptides. Insets show photos of the gel, solution and suspension, respectively. (d) a-proton region of the DOSY spectra for KYF, KFD and GGG tripeptides at 10 mM at pH 7. Horizontal lines indicate the value for the logarithm of the diffusion constant D. Note that PFF is not included due to its low solubility. (e) FTIR absorption spectra in the amide I region of tripeptides (30 mM in D20 at pH 7): KYF, PFF (and KFD and GGG. Spectra have been vertically offset for clarity. (f) DLS auto-correlation functions for KYF, KFD, PFF and GGG (10 mM at pH 7). The inset shows the hydrodynamic radii (RH in nm) and AP scores at 50 ns for the four peptides.

[0086] The FIG. 1(a) is a system for use in identifying peptides, in accordance with the invention. The system includes a simulation unit 3 and a peptide identification unit 5. In the present embodiment, the system 1 comprises a high-performance computer system. In other embodiments, the computing apparatus may comprise any appropriate computing apparatus or combination of computer apparatuses, for example a PC, a workstation, a mainframe, or a network comprising multiple computer apparatuses.

[0087] The simulation unit 3 is operable to perform a molecular simulation of aggregation for each of a plurality of peptides to determine a measure of propensity for aggregation for each of the peptides.

[0088] The peptide identification unit 5 is operable to adjust the measure of propensity for aggregation for each peptide in dependence on a measure of hydrophilicity for the peptide, and to identify at least one peptide in dependence on the hydrophilicity-adjusted aggregation propensity.

[0089] The system 1 may have a user interface 9 (e.g. a keyboard and/or other user input device, one or more screens), a data store 7 in the form of one or more volatile and/or non-volatile data storage devices (e.g. a hard drive, RAM), functioning as a data store, and a processor 2, functioning as the processing resource.

[0090] In the present embodiment, the simulation unit and peptide identification unit are each implemented in the processor 2 by means of a computer program having computer-readable instructions that are executable to perform the method of the embodiment, However, in other embodiments, the units may be implemented in software, hardware, or any suitable combination of software and hardware. Although particular units are described in this embodiment, in alternative embodiments functionality of one or more of these units can be provided by a single unit, processing resource or other component, or functionality provided by a single unit can be provided by two or more units or other components in combination.

[0091] In use, the system 1 functions according to the generalised work flow set out in FIG. 1(b). It will be appreciated by the skilled addressee that the sequence of steps may vary from the sequence set out in FIG. 1(b) and described herein.

[0092] At stage 10, the simulation unit 3 performs a molecular dynamics simulation of each of a plurality of peptides. In some embodiments, the molecular dynamics simulation is performed using GROMACS with the MARTINI force field—Gromacs is a software package for running molecular dynamics simulations [http://www.gromacs.org/About_Gromacs]. In other embodiments, any suitable molecular dynamics method may be used. For each peptide, the simulation unit 3 determines a solvent accessible surface area at the start of the molecular dynamics simulation (SASA.sub.initial) and a solvent accessible surface area at the end of the molecular dynamics simulation (SASA.sub.final). In some embodiments, the determination of SASA is performed using Visual Molecular Dynamics (VMD). In other embodiments, any suitable program for determining SASA may be used, many different programs can be used to calculate the SASA, such as NACCESS—[http://www.bioinf.manchester.ac.uk/naccess/, pymol, etc.]. The simulation unit 3 determines a measure of propensity for aggregation (AP) for each peptide from the determined SASA.sub.Initial and SASA.sub.final using equation 1.

[0093] At stage 20, the peptide identification unit 5 calculates a hydrophilicity-adjusted measure of propensity for aggregation (AP.sub.H) for each peptide. For each peptide, the peptide identification unit 5 retrieves from data store 7 a measure of hydrophilicity for the peptide. The peptide identification unit 5 uses the AP determined at stage 10 and the retrieved measure of hydrophilicity to calculate AP.sub.H using equation 2. Equation 2 is an empirically founded equation.

AP.sub.H=(AP′).sup.α.Math.logP′  (2) [0094] where logP is the measure of hydrophilicity and where the apostrophe indicates the normalization of the respective variable between 0 and 1. The normalisation is determined by scaling all values relative to the range covered by the set of peptides using the following equation:

[00001]${\mathrm{AP}}_H={\left(\frac{\mathrm{AP}.Math..Math.\left(\mathrm{trip}\right)-{\mathrm{AP}}_{\min }}{{\mathrm{AP}}_{\max }-{\mathrm{AP}}_{\min }}\right)}^{\alpha }.Math.\left(\frac{\log .Math..Math.P.Math..Math.\left(\mathrm{trip}\right)-{\log .Math.P}_{\min }}{{\log .Math.P}_{\max }-{\log .Math.P}_{\min }}\right)$

[0095] In equation 2, α is an arbitrary coefficient that can be used to determine the weight of the normalized AP score to the AP.sub.H value. In this embodiment, α=2 was used to give a compromise between AP and hydrophilicity. Depending on the desired properties, the exponent can be decreased (increased) to include more (less) hydrophilic peptides. For a selection with α=1 and α=3 see FIG. 2(d) and (e).

[0096] Any appropriate measure of hydrophilicity may be used, in particular those that relate the hydrophobicity of amino acids to rank the relative hydrophobicity/hydrophilicty of the peptide.

[0097] At stage 30, the peptide identification unit 5 identifies a subset of the plurality of peptides based on the determined AP.sub.H values. In some embodiments, the peptide identification unit 5 identifies a subset by determining the size of the desired subset, ranking the peptides in order of AP.sub.H, and choosing the appropriate number of peptides. For example the peptide identification unit 5 identifies the top 200 peptides in order of AP.sub.H. The peptide identification unit 5 may choose the size of the subset to be a proportion of the entire set of peptides under consideration, for example the top 10% of peptides in order of AP.sub.H.

[0098] In other embodiments, the peptide identification unit 5 determines a threshold value of AP.sub.H. In some embodiments, the threshold value of AP.sub.H is determined from the determined values of AP.sub.H for the peptides, for example by setting a threshold that separates the top 10% of peptides from the remaining peptides. In other embodiments, the threshold value of AP.sub.H may be set by another method, for example by using the determined AP.sub.H of peptides that are already known to have good aggregation properties. In some embodiments, the threshold value of AP.sub.H may be set by a user. The peptide identification unit 5 may identify a subset by identifying all peptides with an AP.sub.H that meets or exceeds the threshold value.

[0099] At stage 40, at least one of the subset of peptides is synthesised using methods described below. At stage 50, the synthesised peptide is allowed to aggregate, for example to form a nanostructure.

[0100] Although the above embodiments described the identification of a subset of peptides in a plurality of peptides, in other embodiments only one peptide is used in the method of the flowchart of FIG. 1(b), with a single determination of AP.sub.H being performed. In such embodiments, a threshold AP.sub.H may be used at stage 30 and the peptide may be identified as potentially suitable for aggregation in solution if its AP.sub.H value exceeds the threshold value.

[0101] Experimental Section

[0102] In an experiment, simulations were performed for all 8000 tripeptides. Tripeptide coordinate files were created using VMD scripting tools.sup.1* and converted to CG representation in the MARTINI force field (version 2.2) using martinize.py with flag −ss =EEE..sup.2 Using the GROMACS code version 4.5.3,3 a cubic box of 13×13×13 nm.sup.3 containing 300 zwitterionic tripeptides was created giving a peptide concentration of 0.23 mol L.sup.−1 in standard CG water, with side chains in their most prevalent charge state at pH 7. Note that the computational concentration is one order of magnitude higher than the experimental one to accelerate the assembly process and thus decrease computational cost. Periodic boundary conditions were used. The box was energy minimized for 5000 steps or until forces on atoms converged to under 200 N. The minimized box was subsequently equilibrated for 500,000 steps of 25 fs, using the Berendsen algorithms.sup.4 to keep temperature (.sub.TT=1 ps) and pressure (.sub.TP=3 ps) around 303 K and 1 bar, respectively. Bond lengths in aromatic side chains and the backbone-side chain bonds in I, V and Y were constrained using the LINCS algorithm..sup.5 The total simulation for this initial screening phase equates 12.5 ns, but this equates to roughly 50 ns ‘effective time’, due to the smoothness of the CG potentials..sup.6,7 All times reported in this paper take into account this speedup factor. For the tripeptides selected for further study, the water in the solvated energy-minimized box was converted to polarizable water (PW).sup.8 to better account for charge screening. This system was then energy-minimized again and run in the NPT ensemble for 4.106 steps, or 400 ns effective time. Finally, for the systems with experimental information available, a similar, but larger simulation was carried out for 1200 ns using 1200 peptides in a box of 24×24×24 nm.sup.3 (peptide concentration 0.14 M in standard CG water) to allow the formation of larger structural features like tubes, spheres or fibers where appropriate.

[0103] The measure of hydrophilicity for a given tripeptide may be written as logP(trip). Since logP(trip) is a unitless number linearly proportional to AGwater-oct (see SI) and is normalized in Eq. 1, it was chosen to define it simply as the sum of the Wimley-White whole-residue hydrophobicities.sup.37,38ΔG.sub.water-octanol (kcal/mol) for the tripeptide, given by:

logP=Σ.sub.i=1.sup.3ΔG.sub.water-octanol   (3)

[0104] Explanation of relationship of hydrophobicity “logP” and ΔG.sub.water-octanol

[0105] Considering the equilibrium

[0106] Amino acid (water)⇄Amino acid (octanol)

[0107] An equilibrium constant K.sub.eq, in this case defined as the partition coefficient P can be written up as

[00002]$K_{\mathrm{eq}}=\frac{{\left[\mathrm{Amino}-\mathrm{acid}\right]}_{\mathrm{oct}}}{{\left[\mathrm{Amino}-\mathrm{acid}\right]}_{\mathrm{wat}}}\equiv P,$

[0108] In chemical equilibrium, the standard Gibbs free energy change AG° can be written as

ΔG.sup.o=−RTlnK.sub.eq

[0109] Showing that the standard free energy change is linearly proportional to the logarithm of P

[00003]$\Delta .Math..Math.G.Math.°=\frac{-\mathrm{RT}}{\log .Math.e}.Math..Math.\log .Math.P$

[0110] AP.sub.H values were calculated for all 8000 tripeptides from 50 ns simulations. A subset of tripeptides was chosen based on the calculated AP.sub.H values, which comprised the 400 tripeptides having the highest AP.sub.H values. An extended molecular dynamics simulation was performed on each of the subset of tripeptides. Computational power allowed only the top 400 scoring peptides indicated after 50 ns by this protocol to be used in extended simulations. A number of the top scoring peptides were synthesised.

[0111] Peptide synthesis

[0112] GGG was purchased from Sigma Aldrich. All other tripeptides were synthesized using standard Fmoc solid phase peptide synthesis with HBTU activation. Briefly, Fmoc Wang resin loaded with the first amino acid was allowed to swell in DMF for 15 min (repeated twice). Fmoc deprotection was performed in 20% piperidine in DMF (repeated twice, for 5 min and 15 min). HBTU activation was performed with 3 eq. of Fmoc-amino acid and 3 eq. of HBTU in DMF (5 ml for every gram of resin), with 6 eq. DIPEA. Coupling was performed at RT for 3 h, after thorough washes with DMF and DCM. Final cleavage was obtained using a mixture of TFA/TIPS/water (95:2.5:2.5). The process was repeated to couple the third amino acid. Resulting peptides were allowed to be precipitated in cold ether or lyophilized when appropriate. The precipitated peptides were allowed to dry under vacuum overnight. The purity of each peptide was verified by HPLC. The peptides' identity was verified by ESI-MS and .sup.1H NMR.

[0113] HPLC

[0114] A 30 μl sample was injected onto a Macherey-Nagel C18 column (Intersil Phenyl for KFD and KHD) with a length of 250 mm and an internal diameter of 4.6 mm and 5-mm fused silica particles at a flow rate of 1 ml/min. The eluting solvent system (in all cases with 0.1% v/v TFA added) had a linear gradient of 20% (v/v) acetonitrile in water for 4 min, gradually rising to 80% (v/v) acetonitrile in water at 35 min. This concentration was kept constant until 40 min when the gradient was decreased to 20% (v/v) acetonitrile in water at 42 min. The purity of each peak was determined by UV detection at 214 nm.

[0115] FTIR

[0116] Samples were contained in a standard IR transmission cell (Harrick Scientific) between two 2 mm CaF2 windows, separated by a polytetrafluoroethylene (PFTE) spacer of 50 μm thickness. Spectra were recorded on a Bruker Vertex 70 spectrometer by averaging 25 scans at a spectral resolution of 1 cm.sup.−1. Spectra were corrected for absorption from a phosphate buffer blank sample and absorptions from trifluoroacetic acid (TFA).

[0117] TEM

[0118] Carbon-coated copper grids (200 mesh) were glow discharged in air for 30 s. The support film was touched onto the gel surface for 3 s and blotted down using filter paper. Negative stain (20 ml, 1% aqueous methylamine vanadate (Nanovan; Nanoprobes) was applied and the mixture was blotted again using filter paper to remove excess. The dried specimens were then imaged using a LEO 912 energy filtering transmission electron microscope operating at 120 kV fitted with 14 bit/2 K

[0119] Proscan CCD Camera

[0120] Diffusion Ordered NMR Spectroscopy (DOSY NMR)

[0121] Peptide solutions were prepared in D.sub.2O at 10 mM peptide concentration as a compromise between solubility and NMR signal intensity. DOSY spectra were acquired at 600 MHz using a Bruker Avance 600 spectrometer using the Bruker microprogram dstebpgp3sat at 298 K. The eddy current delay (Te) was set to 5 ms. The diffusion time was adjusted to 100 ms. The duration of the pulse field gradient, Ag, was optimized in order to obtain 5% residual signal with the maximum gradient strength with the resulting A value of 3.6 ms. The pulse gradient was increased from 2 to 95% of the maximum gradient strength using a linear ramp 16 k data points in the F2 dimension (20 ppm) and 16 data points in the F1 dimension were collected. Final data sizes were 16 k×128.

[0122] Dynamic Light Scattering (DLS)

[0123] The dynamic light scattering (DLS) measurements were carried out on 10 mmol/L peptide solutions (i.e. below the critical gelation concentration) by using a 3 DDLS spectrophotometer (LS instruments, Fribourg, Switzerland) using vertically polarized He-Ne laser light (25 mW with wavelength of 632.8 nm) with an avalanche photodiode detector at an angle of 90° at 25° C. intensity autocorrelation functions were recorded and analyzed by means of the cumulant method in order to determine the intensity weighted diffusion coefficients D and the average hydrodynamic radius R.sub.h by using the Stokes-Einstein equation, R.sub.h=k.sub.BT/6πηD, where k.sub.B is the Boltzmann constant, T is the absolute temperature and n is the solvent viscosity at the given temperature.

[0124] Preparation of Tripeptide Samples

[0125] Tripeptides were dissolved in water by sonication and vortexing, followed by bringing the pH to 7.2±0.1 by dropwise addition of 0.25 M NaOH solution to a final concentration of 30 mmol/L. These were diluted to 10 mmol/L for DLS and NMR experiments. Note that the KYF gel could also be prepared by direct dissolution into a 0.1 M phosphate buffer at pH 7. Characterization of the resulting samples took places at least 24 hours after sample preparation. Samples for FTIR and NMR analysis were prepared using D2O (99.9% D) and NaOD solution instead of H2O 2O and NaOH.

[0126] In the current work, we provide a design-oriented approach by screening all possible combinations of amino acids in tripeptides (20.sup.3=8000 different sequences) and subsequently identifying those with the best predicted properties for experimental investigation.

[0127] We have chosen to utilize the MARTINI force field, which has been extensively parameterized for amino acids (see FIG. 2(a))..sup.33-35 This force field provides a speed up compared to atomistic force fields by approximately 3 orders of magnitude, allowing access to a sufficient simulation size and length for our new screening approach. The ranking of the output of molecular simulations according to descriptors such as propensity to aggregate and hydrophilicity allows the selection of a set of candidates to form nanostructures in water.

[0128] The following description provides a simulation protocol for rapidly selecting candidates with self-assembly propensity is presented, which is subsequently adapted to favour candidates that include hydrophilic residues. The results of these simulations are analyzed in terms of aggregation propensity, structural features and general design rules and subsequently verified against the experimental results from literature. Finally, a number of new peptides with promising scores were selected from the simulation results, synthesized and analyzed with regards to their assembly behavior by diffusion ordered NMR (DOSY NMR), atomic force and transmission electron microscopy (AFM/TEM), fourier transform infrared (FTIR) spectroscopy and dynamic light scattering (DLS).

[0129] Results and Discussion

[0130] Initial Screening Phase

[0131] 50 ns simulations were performed for all 8,000 tripeptides studied. From the last frame of these simulations, the Aggregation Propensity (AP, see Experimental Section) was measured. A list of high (AP>2) scoring peptides are displayed in FIGS. 3(d) and 3(e) (analogous to FIG. 3(a)). On average, hydrophobic tripeptides have higher AP scores and W, F, Y give rise to relatively high contribution to the AP. Remarkably, T and S also stand out as contributing strongly compared to other amino acids with similar hydrophobicity. However, a wide range of AP scores were observed with intermediately hydrophilic peptides as shown in FIG. 2(c), which displays the AP scores as a function of the total hydrophobicity (logP) of the tripeptide. It is noted that only a weak correlation exists between the total hydrophilicity of the tripeptide and its propensity to aggregate: while hydrophobic peptides (logP>3) always display a relatively high score and hydrophilic peptides (logP<7) have a low tendency to aggregate, intermediately hydrophilic or amphiphilic peptides exhibit a wide range of AP scores from ˜1 (no aggregation) to ˜2.4 (strong aggregation). This confirms that aggregation is not a process that can be predicted solely on basis of a peptide's hydrophilicity and the simulations presented here are needed to distinguish between good and poor candidates for self-assembly.

[0132] Strongly hydrophobic peptides are often insoluble in water. However, our screening approach does not, a priori, exclude any peptide based on known practical solubility limitations. As such, hydrophobic peptides produce high AP scores. However, the simulations presented here are of insufficient size and detail to distinguish between the processes of aggregation, precipitation, crystallization or self-assembly. Therefore, in order to select a more appropriate subset of peptides for practical use we have developed a hydrophilicity-corrected score, AP.sub.H, which introduces a positive bias towards hydrophilic peptides (for details see experimental methods). The inclusion of hydrophilic residues has previously been shown to transform insoluble nanofibers to a hydrogel network..sup.36

[0133] FIG. 3(a) shows the normalized APH score for all 8000 tripeptides. The amino acids on the axes are ordered from hydrophobic to hydrophilic according to the Wimley-White scale..sup.37,38 It is clear that a different set of peptides is indicated as having self-assembly propensity combined with hydrophilicity when compared to FIG. 3(d). Contributions from peptides containing charged (K, R, D, E) and hydrogen bonding amino acids (mainly T and S) stand out more using the AP.sub.H score, which will be further discussed in the next section. The top 400 peptides selected by this protocol are indicated in FIG. 2(c) and Table 3 and were subjected to an extended screening phase.

[0134] Generation of Design Rules

[0135] The simulation results of all 8000 tripeptides were analyzed by studying the contributions of specific amino acids to the AP.sub.H score. This allows for the determination of design rules, i.e., placement of certain amino acids in a particular position within the peptide chain to promote self-assembly. FIG. 3(c) shows the average AP.sub.H score of all peptides with a certain amino acid in position 1 (N-terminus), 2 or 3 (C-terminus). Several interesting observations were made: (I) Aromatic amino acids (F, Y and to a lesser extent W) are more favorable in position 2 and 3 compared to the N-terminal amino acid, (II) negatively charged amino acids (E and especially D) are strongly favored in position 3 and (III) positively charged and hydrogen bond donating amino acids (K, R; S, T) promote self-assembly when located at the N-terminus. Also proline is favored in position 1, which could be due to its unique conformational properties allowing better packing of the short peptides; the ‘kink’ in the backbone chain leading to more ordered self-assembly. These design rules were used to select a number of tripeptides from across the range of AP and AP.sub.H scores to study experimentally (see Table 1 and below).

[0136] Evaluation of Simulation Length

[0137] The emergence of structural features in peptide self-assembly takes place on timescales that are longer than the 50 ns used in the initial screening phase..sup.39 To test the validity of using the results of the initial 50 ns phase, the simulation time was extended to 400 ns for peptides with either a high AP.sub.H score (top 5% of 8000) or with an AP score>2 (124 peptides, 1.55%, including 53 peptides that fall in both categories), as highlighted in FIG. 2(c). This cut-off has been previously shown to be reasonable for the AP score for self-assembly candidates in the study of all dipeptides..sup.28 Moreover, for these simulations the standard coarse-grain water was replaced by “polarizable” coarse-grain beads, which have been shown to represent the polarizability and charge screening of water more accurately.

[0138] Table 1 shows the AP scores of the experimentally studied peptides at 50 and 400 ns. When comparing the values at 50 and 400 ns it becomes apparent that AP scores do increase slightly when extending the simulation, but no peptides were observed to change from ‘not aggregating’ to ‘strongly aggregating’ or vice versa and their relative scores remain broadly constant. Overall this data suggests that standard CG water and 50 ns simulation length is sufficient for an initial screening phase for these systems.

[0139] Evaluation of Examples From the Literature.

[0140] The top half of Table 2 shows all tripeptides that were examined for their assembling properties found in literature. CFF, FFF, LFF, VFF and FFV were reported to form extended nanostructures under mainly aqueous conditions..sup.8-11 All these tripeptides were found to have an AP score>2 after the 50 ns simulation, which supports that this is a reasonable cut-off for assembling peptides. Interestingly, Marchesan et al. noted that VFF (AP=2.3) showed more evidence for nanostructure formation than structural isomer FFV (AP=2.0),.sup.10 which agrees with our observation that aromatics in the 2.sup.nd and 3.sup.rd position improve aggregation. KFG, reported to form vesicles and nanotubes has an AP of 1.6, but does largely obey our design rules with K at the N-terminus and F in the second position.

[0141] To study the formation of nanostructures in the simulations, the simulation time was extended even further on a larger peptide box (1200 peptides, 0.14 M in CG water) for the tripeptides studied experimentally in literature. The last frames of the MD simulations are displayed in Table 2 and are in agreement with the experimentally observed nanostructures in Table 2: FFF (experiment: spheres, plates), CFF, VYV (experiment: spheres) and FFV and LFF (experiment: fibers) match the computational result. Longer simulations on FFF were reported to produce nanospheres and nanorods, similar to the structures observed here..sup.31 Tripeptides DFN and ECG, which were experimentally not found to exhibit assembly in solution,.sup.15,41 were also calculated to have a low propensity to aggregate. It should be noted that the size of the aggregates in these simulations does not compare to the experimental size of the aggregates due to the limited number of molecules in the simulation. It has become apparent the proposed computational method is mainly valuable in determining a good set of candidates for self-assembling nanostructures after 50 ns, but structural information can be obtained as well with much longer, extended, simulations as was reported previously..sup.28,30,31

[0142] Model Predictive Value

[0143] The computational method outlined above was validated by comparison with experimental work from literature (vide supra). However, the added value of the proposed procedure lies in identifying new self-assembling peptides. We tested its predictive value by synthesizing and characterizing various tripeptides indicated as good candidates for nanostructure formation.

[0144] The bottom half of Table 2 contains the tripeptides that were selected for synthesis and experimental characterization. The selection was made to include a wide range of AP and AP.sub.H scores, including PFF (#1 from AP score) and KFD (#1 from AP.sub.H score) and peptides that adhere to one or more of our design rules formulated above, such as KYF. GGG is included as a non-self-assembling control peptide (AP=1.0). These peptides were dissolved in water at pH 7 and their aggregation behavior was studied in order to test the predictive value of our method. For KYF, this afforded the formation of a translucent self-supporting hydrogel, while PFF and FHF gave suspensions, IYF and FYI precipitated and the other tripeptides gave clear solutions.

[0145] To correlate the AP scores calculated with experimentally observed aggregation, the presence and size of aggregates were determined by Diffusion Ordered NMR spectroscopy (DOSY). This method is able to determine the diffusion coefficient D of supramolecular aggregates and therefore their relative size, as described by the Stokes-Einstein equation..sup.43,44 Diffusion constants for all the studied tripeptides can be found in Table 2. FIG. 4(d) displays the a-proton region of the DOSY spectra of KYF, KFD and GGG. The poor solubility of PFF, FHF, IYF and FYI prevented reliable extraction of aggregate size for these peptides. For the remaining peptides, DOSY results are in good agreement with the predicted aggregation propensity at 50 ns. As expected, the negative control GGG has the highest diffusion constant. KFD has a relatively low AP and shows an intermediate diffusion rate. KYF has the lowest diffusion constant, consistent with its tendency to form an aggregated hydrogel structure. It can be seen from FIGS. 2(c) and Table 2 that KYF ranks high on the AP.sub.H score (#28 out of 8000), again demonstrating the value of taking the hydrophilicity of peptides into account (KYF only ranks #478 in the AP score). Note that peptide RYF (#212, AP.sub.H score) did not give a gel and SYF (#380, AP.sub.H score) was insoluble in water (data not shown).

[0146] Further studies were employed to elucidate the nanostructural components of KYF (gel), KFD (#1 from AP.sub.H score), PFF (#1 from AP score) and GGG (negative control) specifically, using FTIR spectroscopy, DLS and TEM.

[0147] The secondary structure of the peptides can be probed by infrared (IR) spectroscopy..sup.45,46 When aggregation takes place via intermolecular hydrogen bonding of the amide groups, the 1650-1655 cm.sup.−1 absorption observed for free peptides in solution typically narrows and shifts to lower frequency, while the 1595 cm.sup.−1 peak, assigned to carboxylate groups, broadens or decreases in intensity due to protonation or salt bridge formation. This was clearly observed for PFF in FIG. 3(e), as the amide I absorption shifts to 1637 cm.sup.−1 and significantly narrows compared to non-aggregating peptide GGG, consistent with a n-sheet like arrangement of the amide groups. KFD shows no significant narrowing of the amide absorption compared to GGG, indicating no extended secondary structure formation. Note that small absorption maxima in the 1670-1685 cm.sup.−1 region are assigned to residual trifluoro-acetic acid absorption and the 1570-1580 cm.sup.−1 band for KFD is assigned to the aspartic acid side chain carboxylate group.

[0148] For KYF, a dramatic change in the amide region was noticed upon gelation with intense IR absorption peaks at 1621 and 1649 cm.sup.−1 for the amide groups and 1568 cm.sup.−1 for salt-bridged carboxylate groups. This indicates strong intramolecular hydrogen bonding between amide modes and strong interactions of the N-terminus or lysine side chains with the C-terminus, both suggesting a well-ordered peptide nanostructure..sup.45,46. To further confirm the presence (and size) of the peptide aggregates formed, dynamic light scattering (DLS) experiments were performed on the KYF gel, PFF suspension and KFD and GGG solutions. FIG. 3(h) shows the intensity auto-correlation function and the average hydrodynamic radius of the aggregates determined from the extracted diffusion constant. The relative size of the aggregates (0.26, 0.13, 0.70 and 0.10 μm for KYF, KFD, PFF and GGG, respectively) is in reasonable agreement with the relative AP scores for these peptides as predicted by the MD simulations (Table 2), although it was surprising to see such large aggregates for GGG. For the peptides synthesized here, TEM studies on KYF, PFF and KFD reveal an entangled fibrous network for the KYF hydrogel (FIG. 4(a)). Although KYF was not observed to form fibers after 1200 ns in the MD simulation, further extension of the simulation to 4800 ns resulted in a branched fibrous nanostructure spanning the whole simulation box. TEM images of PFF (FIG. 4(b)) reveal short crystalline nanostructures with a large aspect ratio, while KFD (FIG. 4(c)) was observed to form strongly curved fibers together with more amorphous regions.

[0149] Investigation into the pH Dependence on the Gelation Behaviour of the KYF Hydrogelator

[0150] We investigated the effects of pH on the gelation of the tripeptide KYF in aqueous solvent. Infrared spectroscopy of the KYF gel shows that the formation of a salt bridge plays a significant role in the self-assembly of KYF. The alteration of the pH will affect the protonation state of the ionisable groups on KYF i.e., N and C termini and the lysine side groups. Herein, the inventors demonstrate that by changing the pH of an aqueous solution of KYF, gelation can be induced.

[0151] pH measurements were carried out ranging from pH 5 to 8. At each pH measurement an image was taken to show the gelation behaviour. The pH was altered using 0.5M NaOH which is added drop wise into the vial, vortexed to ensure a complete mix then the pH is measured until a stable value is obtained.

[0152] Initially, the peptide sample is dissolve in water and the pH is recorded. At this point the pH is below the pKa of the C-terminus therefore the terminal acid is protonated and gelation should not occur, which is what was observed.

[0153] On changing the pH of the sample from 2.21 to 6.7 similar results were obtained. Namely, these samples remained as a clear solution and the gel state was not observed. However, as the pH is increased to 6.9 a weak gel is formed. A tilted vial showed the presence of a gel material but on moving the vial the gel collapsed and water began to separate out. At pH 7.05 a stronger gel is formed. At this point the vial can be rotated fully showing a self-supporting hydrogel. On increasing the pH from 7.05 until 7.98, we observed the formation of stronger gels. Moreover, it was noted that there is an increase in turbidity of the gel, i.e., on increasing the pH the gel become cloudier.

CONCLUSIONS

[0154] We have presented a coarse-grain MD protocol for screening peptides for their aggregation behavior and applied this to the set of 8000 gene-encoded tripeptides. After an initial 50 ns screening phase a subset of peptides was selected based on their aggregation propensity (AP) and hydrophilicity-corrected AP (AP.sub.H). This set was then used in extended simulations to study the tripeptide dynamics and nanostructure formation.

[0155] The simulation results indicate only a weak correlation between hydrophilicity and AP, confirming that self-assembly propensity is not simply a measure of hydrophobicity and clearly illustrating the usefulness of MD simulations for selection of self-assembling peptides. Furthermore, a set of design rules that promote aggregation was described, where aromatic amino acids are most favorable in positions 2 and 3 in a tripeptide, while positive and H-bonding residues favor position 1 (N-terminus) and negative residues position 3 (C-terminus).

[0156] The results of the simulations were validated by comparison with experimental results from literature and by synthesis and characterization of a set of tripeptides which are indicated by our method and design rules to be promising candidates for self-assembly.

[0157] Interestingly, this led to the discovery of the first unprotected full-L tripeptide (KYF) that forms a hydrogel in the absence of organic solvents. More generally, excellent agreement was observed between predicted AP scores and experimental behavior.

[0158] These results support further exploitation of the protocol for screening larger peptides in the search for new nanostructures.

[0159] Finally, in this current work we have presented a selection method based on the hydrophilicity of the tripeptides. However, it is clear that depending on the desired properties of the self-assembling peptide filters can be added, such as automated screening based on the shape of nanostructure,.sup.47 peptide foldability,.sup.48 the presence of certain amino acid interactions or other molecular properties, such as protonation state of the amino acids alone and/or when part of a peptide sequence could be equally useful and constitute further additions.

[0160] In the pH experiments, it is noted that the critical gelation occurs at a pH that is very close to neutral (6.9-7.05). This result was unexpected based on the pKa values of the ionisable groups (C-terminal, N-terminal, and lysine side-chain) in the peptide. Clearly, the hydrophobic environment that is created upon assembly of the KYF peptide causes a tunable shift in pKa (tunable by hydrophobicity of the overall peptide) of one or more of its ionisable moieties.

[0161] The pH dependence on the gelation of KYF has been demonstrated. We have seen that at neutral pH ranges hydrogelation occurs, but at lower pH ranges no gelation is observed. There is clearly a relationship between the ability of the tripeptide to form a nanofibrous network that is capable of forming a gel and the protonation states of the amino acids that compose the tripeptide. This principle could be more generally extended to other tripeptides that we have shown to form hydrogels, or that could be designed to form hydrogels following the rules described herein. In particular, those peptides that score high on the AP.sub.H score are candidates for pH dependent self-assembly.

[0162] Finally, the screening approach that is described herein can be adopted to screen for the ability of the tripeptides to self-assemble within various protonation states by altering the bead types that describe the amino acids within the coarse grained model. The beads that describe the groups of atoms within the amino acid are parameterised to represent the vdW parameters and charge states of the atoms that comprise the bead. Therefore, by changing the bead from a positive (negative) to neutral bead type, or vice versa, with the associated change in the vdW parameters, allows us to model the different protonation states of the amino acids that compose the tripeptide. This approach would add an additional filter to the screening process described herein, whereby the aggregation propensity (AP), the hydrophilicity and the protonation states of the amino acids would all be analysed to determine whether a tripeptide will be able to form a pH dependent gel state.

TABLE-US-00001 TABLE 1 AP scores after 50, 400 and 1200 ns. AP AP AP AP Trip. 50 ns.sup.a 50 ns.sup.b 400 ns.sup.b 1200 ns.sup.c CFF 2.05 2.21 3.04 3.70 DFN 1.17 1.10 1.11 1.28 ECG 1.01 1.02 0.99 1.05 FFF 2.26 2.44 2.42 3.44 FFV 2.03 2.21 2.84 3.91 KFG 1.56 1.39 1.57 1.94 LFF 2.07 2.41 2.70 3.34 VFF 2.33 2.24 2.92 3.82 VYV 1.88 1.78 2.44 2.67 PFF 2.39 2.20 3.51 3.70 FYI 2.22 2.12 2.24 3.19 FHF 2.09 1.95 3.22 3.17 YFI 1.97 1.93 2.15 3.11 IYF 1.89 2.26 2.18 3.24 KYF 1.85 1.70 1.83 3.19 KFD 1.73 1.40 1.70 2.33 KHD 1.63 1.44 1.92 2.51 KFF 1.92 1.72 2.01 2.66 KYY 1.79 1.58 1.88 3.03 KYW 1.78 1.76 1.99 2.87 KLL 1.31 1.06 1.08 1.38 FYK 1.73 1.63 1.67 2.59 RYF 1.80 1.60 1.98 2.66 GGG 1.07 1.07 1.04 1.11 .sup.a300 peptides in standard CG water, 0.23M, .sup.b300 peptides in polarizable CG water, 0.23M, .sup.c1200 peptides in standard CG water, 0.14M

TABLE-US-00002 TABLE 3 Table S5: peptides from the top 400 AP.sub.H pep AP logP AP_H KFD 1.73 4.73 0.187 KWD 1.74 4.35 0.186 HKD 1.65 6.55 0.176 PFF 2.39 −3.28 0.174 KWE 1.72 4.34 0.174 WKE 1.71 4.34 0.169 KHD 1.63 6.55 0.167 PCF 2.08 −1.59 0.167 KWF 2.00 −1.00 0.163 KFW 2.00 −1.00 0.163 KHE 1.62 6.54 0.161 TSF 1.99 −1.00 0.161 SCW 2.06 −1.65 0.161 WKD 1.69 4.35 0.160 KYD 1.64 5.73 0.160 KEH 1.62 6.54 0.159 GFF 2.14 −2.27 0.158 VAW 2.10 −2.05 0.158 SSF 1.96 −0.79 0.155 KYE 1.63 5.72 0.155 STF 1.97 −1.00 0.154 SRD 1.62 5.91 0.154 KYY 1.79 1.38 0.152 PVF 2.08 −2.03 0.152 SKD 1.59 6.90 0.152 TFP 2.00 −1.32 0.152 RHD 1.63 5.56 0.151 KYF 1.85 0.38 0.151 FKF 1.92 −0.62 0.150 KFF 1.92 −0.62 0.150 WFD 1.88 −0.16 0.149 WRD 1.69 3.36 0.149 FKD 1.65 4.73 0.148 PHF 2.00 −1.46 0.148 KFY 1.84 0.38 0.147 KFE 1.64 4.72 0.147 PPF 1.99 −1.43 0.147 SFE 1.73 2.38 0.147 SNH 1.78 1.42 0.147 WKF 1.94 −1.00 0.146 CPW 2.05 −1.97 0.145 SYS 1.84 0.21 0.145 VPF 2.05 −2.03 0.144 PYY 1.96 −1.28 0.143 YFK 1.83 0.38 0.143 TKD 1.58 6.69 0.143 FKY 1.82 0.38 0.142 KDF 1.63 4.73 0.141 LCF 2.18 −2.98 0.140 TCW 2.01 −1.86 0.140 GFY 1.95 −1.27 0.140 TGF 1.87 −0.31 0.140 SWE 1.73 2.00 0.139 STY 1.84 0.00 0.139 TYF 2.05 −2.17 0.139 PMW 2.11 −2.62 0.139 PSF 1.93 −1.11 0.139 TFV 2.02 −1.92 0.138 SKW 1.77 1.17 0.138 SFD 1.71 2.39 0.138 SPF 1.93 −1.11 0.138 PGW 1.90 −0.80 0.137 FFD 1.82 0.22 0.137 KFT 1.75 1.34 0.136 PPW 1.99 −1.81 0.136 RWD 1.66 3.36 0.136 TRD 1.59 5.70 0.136 PFT 1.94 −1.32 0.135 FCC 1.98 −1.75 0.135 RFY 1.87 −0.61 0.135 PIY 1.97 −1.69 0.134 RFD 1.64 3.74 0.134 SHK 1.66 3.37 0.134 SHN 1.74 1.42 0.133 YKD 1.58 5.73 0.133 SGF 1.83 −0.10 0.133 HRD 1.58 5.56 0.132 VVF 2.08 −2.63 0.131 TSH 1.76 0.82 0.131 FRY 1.86 −0.61 0.130 TTF 1.91 −1.21 0.130 FKW 1.89 −1.00 0.130 YPV 1.89 −1.03 0.129 PTF 1.91 −1.32 0.128 VFF 2.33 −3.88 0.128 CFV 2.01 −2.19 0.128 SFY 1.98 −1.96 0.128 PCW 1.98 −1.97 0.128 CSY 1.82 −0.27 0.128 YFC 2.04 −2.44 0.128 SWP 1.93 −1.49 0.128 FVV 2.07 −2.63 0.127 SGW 1.84 −0.48 0.127 ISW 2.08 −2.75 0.127 CFC 1.95 −1.75 0.126 IFG 1.94 −1.68 0.126 KVW 1.78 0.25 0.126 YFD 1.73 1.22 0.126 FDF 1.78 0.22 0.126 SYH 1.81 −0.14 0.126 SHF 1.88 −1.14 0.125 VFG 1.87 −1.02 0.125 WRE 1.63 3.35 0.125 KDW 1.60 4.35 0.125 PYV 1.87 −1.03 0.125 TYT 1.81 −0.21 0.124 FYI 2.22 −3.54 0.124 PFY 2.00 −2.28 0.124 YPY 1.89 −1.28 0.124 VFD 1.71 1.47 0.124 PFV 1.97 −2.03 0.124 SFW 2.18 −3.34 0.124 WFG 2.05 −2.65 0.124 TFC 1.91 −1.48 0.123 PFS 1.87 −1.11 0.123 STH 1.74 0.82 0.123 FGS 1.79 −0.10 0.123 VYY 1.95 −1.88 0.123 FRD 1.61 3.74 0.123 VSW 1.97 −2.09 0.122 VFE 1.70 1.46 0.122 KDY 1.55 5.73 0.122 CMF 2.01 −2.40 0.122 PSY 1.79 −0.11 0.122 YGF 1.88 −1.27 0.122 MFY 2.12 −3.09 0.122 MFF 2.36 −4.09 0.122 TFD 1.67 2.18 0.121 SHD 1.59 4.21 0.121 SWK 1.71 1.17 0.121 RDF 1.61 3.74 0.121 IFT 2.03 −2.58 0.121 LPF 2.07 −2.82 0.121 VCF 1.98 −2.19 0.121 LFD 1.74 0.68 0.120 KPF 1.71 1.23 0.120 TYD 1.63 3.18 0.120 RYD 1.58 4.74 0.120 TKF 1.70 1.34 0.120 SVW 1.96 −2.09 0.120 TWC 1.94 −1.86 0.120 CTW 1.94 −1.86 0.120 KWM 1.77 0.04 0.120 GFW 2.04 −2.65 0.120 GFT 1.80 −0.31 0.120 KYW 1.78 0.00 0.120 VAF 1.91 −1.67 0.119 HRE 1.55 5.55 0.119 IYY 2.02 −2.54 0.119 PSW 1.89 −1.49 0.119 VWK 1.76 0.25 0.119 SGY 1.72 0.90 0.119 HKE 1.53 6.54 0.118 AFY 1.94 −1.92 0.118 SYY 1.84 −0.96 0.118 RWE 1.61 3.35 0.118 CYH 1.82 −0.62 0.118 SCY 1.79 −0.27 0.118 CFG 1.81 −0.58 0.118 KEW 1.58 4.34 0.118 KWC 1.73 0.69 0.118 VFP 1.95 −2.03 0.118 SHS 1.71 1.03 0.118 SWH 1.89 −1.52 0.118 PLY 1.92 −1.82 0.117 SKE 1.52 6.89 0.117 TFT 1.86 −1.21 0.117 HKF 1.70 1.20 0.117 FTC 1.88 −1.48 0.117 TYI 1.89 −1.58 0.117 CLW 2.15 −3.36 0.117 FIS 1.98 −2.37 0.116 PWH 1.92 −1.84 0.116 VFK 1.73 0.63 0.116 KLF 1.77 −0.16 0.116 CFD 1.66 1.91 0.116 PWF 2.21 −3.66 0.116 PVW 1.99 −2.41 0.116 CYF 1.99 −2.44 0.116 PCY 1.80 −0.59 0.116 VFC 1.96 −2.19 0.116 TSY 1.76 0.00 0.116 WDF 1.77 −0.16 0.116 FMC 1.98 −2.40 0.116 LPY 1.91 −1.82 0.115 FKE 1.56 4.72 0.115 YFG 1.86 −1.27 0.115 THS 1.71 0.82 0.115 WEF 1.77 −0.17 0.115 YCC 1.81 −0.75 0.115 WSY 1.97 −2.34 0.115 LSF 1.99 −2.50 0.115 KDH 1.52 6.55 0.115 IYC 1.91 −1.85 0.115 FSM 1.92 −1.92 0.115 CPF 1.89 −1.59 0.115 SWT 1.87 −1.38 0.115 KMW 1.76 0.04 0.115 ASW 1.84 −1.13 0.115 FYT 1.95 −2.17 0.115 KIW 1.79 −0.41 0.115 THF 1.86 −1.35 0.114 PKW 1.71 0.85 0.114 MHF 1.96 −2.27 0.114 VWH 1.98 −2.44 0.114 VGW 1.86 −1.40 0.114 FYK 1.73 0.38 0.114 CWP 1.92 −1.97 0.114 KTF 1.68 1.34 0.114 EFW 1.77 −0.17 0.114 FFM 2.31 −4.09 0.114 RYF 1.80 −0.61 0.114 GFI 1.89 −1.68 0.114 FRE 1.59 3.73 0.114 PWV 1.97 −2.41 0.114 CGY 1.73 0.42 0.113 PFA 1.83 −1.07 0.113 SWC 1.89 −1.65 0.113 PFC 1.88 −1.59 0.113 KFH 1.69 1.20 0.113 VFS 1.89 −1.71 0.113 CFY 1.98 −2.44 0.113 VYV 1.88 −1.63 0.113 PTY 1.77 −0.32 0.113 IFP 2.01 −2.69 0.113 SHT 1.71 0.82 0.113 PSH 1.71 0.71 0.113 SCF 1.85 −1.27 0.113 KFV 1.71 0.63 0.112 REY 1.56 4.73 0.112 PHS 1.71 0.71 0.112 FFR 1.88 −1.61 0.112 VYF 2.04 −2.88 0.112 FFE 1.74 0.21 0.112 FSP 1.83 −1.11 0.112 KWT 1.69 0.96 0.112 CWM 2.02 −2.78 0.112 WPP 1.90 −1.81 0.112 YKF 1.73 0.38 0.112 TSW 1.85 −1.38 0.112 RHE 1.53 5.55 0.111 GFL 1.89 −1.81 0.111 KEY 1.53 5.72 0.111 FGF 1.95 −2.27 0.111 FVS 1.88 −1.71 0.111 CTF 1.86 −1.48 0.111 WWK 1.85 −1.38 0.111 KIF 1.75 −0.03 0.111 WFK 1.82 −1.00 0.111 KWL 1.78 −0.54 0.111 FWE 1.76 −0.17 0.111 RFF 1.87 −1.61 0.111 SFG 1.75 −0.10 0.111 SFC 1.84 −1.27 0.111 PHV 1.76 −0.21 0.111 STS 1.68 1.17 0.111 KHW 1.70 0.82 0.111 IGY 1.79 −0.68 0.111 VFH 1.92 −2.06 0.111 KSF 1.66 1.55 0.110 CYS 1.76 −0.27 0.110 SKH 1.59 3.37 0.110 FYP 1.94 −2.28 0.110 YWD 1.70 0.84 0.110 TGS 1.65 1.86 0.110 VKW 1.73 0.25 0.110 TWD 1.65 1.80 0.110 FYD 1.68 1.22 0.110 KFS 1.66 1.55 0.110 RDW 1.59 3.36 0.110 SSY 1.73 0.21 0.110 PYT 1.76 −0.32 0.110 WYT 1.98 −2.55 0.110 PAW 1.85 −1.45 0.109 FYC 1.96 −2.44 0.109 WGT 1.79 −0.69 0.109 VTW 1.94 −2.30 0.109 FGC 1.78 −0.58 0.109 SFF 2.04 −2.96 0.109 KSW 1.67 1.17 0.109 YFM 2.06 −3.09 0.109 WST 1.84 −1.38 0.109 IPY 1.87 −1.69 0.109 TWY 1.97 −2.55 0.109 YYI 1.97 −2.54 0.109 PWT 1.87 −1.70 0.109 WHP 1.89 −1.84 0.109 VKF 1.70 0.63 0.109 PFG 1.76 −0.42 0.109 SVY 1.78 −0.71 0.108 TFF 2.07 −3.17 0.108 FLP 2.01 −2.82 0.108 CYC 1.79 −0.75 0.108 CGF 1.77 −0.58 0.108 RWY 1.81 −0.99 0.108 PWC 1.90 −1.97 0.108 PPY 1.76 −0.43 0.108 PYS 1.74 −0.11 0.108 SWN 1.79 −0.78 0.108 WEK 1.55 4.34 0.108 FHF 2.09 −3.31 0.108 KEF 1.54 4.72 0.108 CKW 1.69 0.69 0.108 RFE 1.57 3.73 0.108 KHF 1.67 1.20 0.108 SYT 1.73 0.00 0.107 TSS 1.67 1.17 0.107 FWD 1.74 −0.16 0.107 TKH 1.59 3.16 0.107 FRF 1.86 −1.61 0.107 SRE 1.51 5.90 0.107 SHP 1.69 0.71 0.107 SGS 1.63 2.07 0.107 SFH 1.81 −1.14 0.107 WKY 1.73 0.00 0.107 FFA 2.02 −2.92 0.107 WFE 1.74 −0.17 0.107 WFT 2.13 −3.55 0.107 YFE 1.66 1.21 0.106 GVF 1.80 −1.02 0.106 CCY 1.78 −0.75 0.106 TFL 1.98 −2.71 0.106 SGH 1.64 1.72 0.106 RYE 1.54 4.73 0.106 IFA 1.93 −2.33 0.106 FGV 1.80 −1.02 0.106 FKS 1.65 1.55 0.106 VWV 2.03 −3.01 0.106 CYW 2.00 −2.82 0.106 YFS 1.89 −1.96 0.106 FSV 1.86 −1.71 0.106 TDW 1.63 1.80 0.105 SIY 1.83 −1.37 0.105 KWW 1.83 −1.38 0.105 FFG 1.92 −2.27 0.105 KIY 1.67 0.97 0.105 FWP 2.15 −3.66 0.105 CSF 1.82 −1.27 0.105 EKW 1.55 4.34 0.105 CWD 1.64 1.53 0.105 FSL 1.95 −2.50 0.105 FTP 1.82 −1.32 0.105 FFC 2.10 −3.44 0.105 TWG 1.77 −0.69 0.105 FTL 1.97 −2.71 0.105 TKY 1.61 2.34 0.104 PWI 2.03 −3.07 0.104 SRW 1.71 0.18 0.104 KFI 1.72 −0.03 0.104 WLT 2.03 −3.09 0.104 SMY 1.78 −0.92 0.104 HDK 1.49 6.55 0.104 THT 1.69 0.61 0.104 KSD 1.48 6.90 0.104 WKT 1.67 0.96 0.104 FYG 1.81 −1.27 0.104 GCW 1.79 −0.96 0.104 SRF 1.69 0.56 0.104 MFD 1.65 1.26 0.104 IKW 1.75 −0.41 0.104 WSS 1.80 −1.17 0.104 VHS 1.71 0.11 0.104 TTH 1.68 0.61 0.104 SYP 1.73 −0.11 0.104 WKS 1.66 1.17 0.104 FGI 1.85 −1.68 0.104 VMW 2.05 −3.22 0.103 TKW 1.67 0.96 0.103 SWA 1.80 −1.13 0.103 SYK 1.60 2.55 0.103 FFT 2.04 −3.17 0.103 SNY 1.68 0.60 0.103 SNF 1.74 −0.40 0.103 PAF 1.79 −1.07 0.103 GWY 1.84 −1.65 0.103 DKH 1.49 6.55 0.103 YCT 1.75 −0.48 0.103 CYT 1.75 −0.48 0.103 FFS 2.00 −2.96 0.103 SYF 1.87 −1.96 0.103 SVF 1.85 −1.71 0.103 WKI 1.74 −0.41 0.103 PIF 1.96 −2.69 0.103 AFC 1.80 −1.23 0.103 TYS 1.72 0.00 0.103 RHS 1.60 2.38 0.103 SYW 1.92 −2.34 0.103 KWH 1.67 0.82 0.103 PYM 1.80 −1.24 0.103 FPG 1.74 −0.42 0.103 STT 1.66 0.96 0.102 TFK 1.64 1.34 0.102 VYT 1.78 −0.92 0.102 FTI 1.95 −2.58 0.102 NSW 1.77 −0.78 0.102 TIY 1.83 −1.58 0.102 SFK 1.63 1.55 0.102 MIF 2.10 −3.50 0.102 CWC 1.89 −2.13 0.102

TABLE-US-00003 TABLE 4 Moments of inertia along x, y and z axes, the number of molecules in the largest cluster, the number of clusters, aspect ratio and observed morphology from the final frames of the 1200 ns MD simulations on 1200 peptides in a box. # mol. I.sub.x (10.sup.6 I.sub.y (10.sup.6 I.sub.z (10.sup.6 in number amu * amu * amu * largest of Aspect Trip nm.sup.2) nm.sup.2) nm.sup.2) cluster clusters ratio (I.sub.z/I.sub.x) Main morphology CFF 3.82 5.69 6.82 791 6 1.79 Spherical DEN 0.02 0.03 0.03 31 453 2.02 N/A ECG 0.00 0.00 0.00 4 1053 2.77 N/A FFF 1.70 1.80 2.40 396 7 1.41 Sphere/plate FFV 2.61 6.38 6.52 752 2 2.49 Oblong FHF 5.14 13.83 14.54 802 8 2.83 Fibrous.sup.a FYI 2.68 6.07 6.30 743 5 2.35 Oblong FYK 1.78 6.14 6.38 483 6 3.59 Fibrous GGG 0.00 0.00 0.00 21 904 3.69 N/A IYF 0.91 1.09 1.19 328 5 1.30 Spherical KFD 2.70 17.80 18.20 561 5 6.75 Fibrous.sup.a KFF 0.98 5.26 5.48 406 7 5.57 Fibrous KFG 0.38 1.15 1.27 182 44 3.37 Small/fibrous KHD 14.66 60.67 69.42 1164 2 4.74 Fibrous.sup.a KLL 0.18 0.84 0.90 132 221 4.93 N/A KYF 4.36 6.60 9.19 623 5 2.11 Fibrous.sup.a KYW 3.15 8.58 10.67 491 5 3.38 Fibrous.sup.a KYY 1.25 2.19 2.58 362 6 2.06 Spherical LFF 4.24 11.67 12.20 1002 3 2.88 Oblong PFF 2.66 4.64 5.17 693 2 1.95 Spherical/oblong RYF 1.34 6.67 6.86 406 8 5.13 Fibrous VFF 2.68 3.85 4.22 664 4 1.57 Spherical VYV 0.71 0.78 0.86 311 10 1.21 Small/spherical YFI 1.24 2.18 2.31 433 8 1.86 Spherical .sup.aDue to branching the aspect ratio becomes unrepresentatively low

REFERENCES

[0163] 1. Zhang, S. G. Fabrication of novel biomaterials through molecular self-assembly. Nat. Biotechnol. 21, 1171-1178 (2003).

[0164] 2. Hartgerink, J. a, Benlash, E. & Stupp, S. L. Self-Assembly and Mineralization of Peptide-Amphiphile Nanofibers. Science 294, 1684-1688 (2001).

[0165] 3. Fletcher, J. M. et al. Self-Assembling Cages from Coiled-Coil Peptide Modules. Science 340, 595-599 (2013).

[0166] 4. O′Leary, L. E. R., Fallas, J. A., Bakota, E. L., Kang, M. K. & Hartgerink, J. D. Multi-hierarchical self-assembly of a collagen mimetic peptide from triple helix to nanofibre and hydrogel. Nat. Chem. 3, 821-828 (2011).

[0167] 5. Reches, M. & Gazit, E. Casting metal nanowires within discrete self-assembled peptide nanotubes. Science 300, 625-627 (2003).

[0168] 6. Ghadiri, M. R., Granja, J. R., Milligan, R. A., McRee, D. E. & Khazanovich, N. Self-assembling organic nanotubes based on a cyclic peptide architecture. Nature 366, 324-327 (1993).

[0169] 7. Zhang, S., Holmes, T., Lockshin, C. & Rich, A. Spontaneous assembly of a self-complementary oligopeptide to form a stable macroscopic membrane. Proc. Natl. Acad. Sci. U. S. A. 90, 3334-3338 (1993).

[0170] 8. Reches, M. & Gazit, E. Formation of Closed-Cage Nanostructures by Self-Assembly of Aromatic Dipeptides. Nano Lett. 4, 581-585 (2004).

[0171] 9. Tamamis, P. et al. Self-Assembly of Phenylalanine Oligopeptides: Insights from Experiments and Simulations. Biophys. J. 96, 5020-5029 (2009).

[0172] 10. Marchesan, S., Easton, C. D., Kushkaki, F., Waddington, L. & Hartley, P. G. Tripeptide self-assembled hydrogels: unexpected twists of chirality. Chem. Commun. 48, 2195-2197 (2012).

[0173] 11. Marchesan, S. et al. Unzipping the role of chirality in nanoscale self-assembly of tripeptide hydrogels. Nanoscale 4, 6752-6760 (2012).

[0174] 12. Marchesan, S. et al. Chirality effects at each amino acid position on tripeptide self-assembly into hydrogel biomaterials. Nanoscale 6, 5172-5180 (2014).

[0175] 13. James, J. & Mandal, A. B. The aggregation of Tyr-Phe dipeptide and Val-Tyr-Val tripeptide in aqueous solution and in the presence of SDS and PEO-PPO-PEO triblock copolymer: Fluorescence spectroscopic studies. J. Colloid Interface Sci. 360, 600-605 (2011).

[0176] 14. Moitra, P., Kumar, K., Kondaiah, P. & Bhattacharya, S. Efficacious Anticancer

[0177] Drug Delivery Mediated by a pH-Sensitive Self-Assembly of a Conserved Tripeptide Derived from Tyrosine Kinase NGF Receptor. Angew. Chem. Int. Ed. 53, 1113-1117 (2014).

[0178] 15. Reches, M., Porat, Y. & Gazit, E. Amyloid Fibril Formation by Pentapeptide and Tetrapeptide Fragments of Human Calcitonin. J. BioL Chem. 277, 35475-35480 (2002).

[0179] 16. Hauser, C. A. E. et at. Natural tri- to hexapeptides self-assemble in water to amyloid beta-type fiber aggregates by unexpected alpha-helical intermediate structures. Proc. Natl. Acad. Sci. U. S. A. 108, 1361-1366 (2011).

[0180] 17. Cao, M., Cao, C., Zhang, L., Xia, D. & Xu, H. Tuning of peptide assembly through force balance adjustment. J. Colloid Interface Sci. 407, 287-295 (2013).

[0181] 18. Zhang, Y., Gu, H., Yang, Z. & Xu, B. Supramolecular Hydrogels Respond to Ligand-Receptor Interaction. J. Am. Chem. Soc. 125, 13680-13681 (2003).

[0182] 19. Smith, A. M. & Ulijn, R. V. Designing peptide based nanomaterials—Chemical Society Reviews (RSC Publishing). Chem. Soc. Rev. 37, 664-675 (2008).

[0183] 20. Das, A. K., Bose, P. P., Drew, M. G. B. & Banerjee, A. The role of protecting groups in the formation of organogels through a nano-fibrillar network formed by self-assembling terminally protected tripeptides. Tetrahedron 63, 7432-7442 (2007).

[0184] 21. Subbalakshmi, C., Manorama, S. V. & Nagaraj, R. Self-assembly of short peptides composed of only aliphatic amino acids and a combination of aromatic and aliphatic amino acids. J. Pept. Sci. 18, 283-292 (2012).

[0185] 22. Yang, Z., Liang, G., Ma, M., Gao, Y. & Xu, B. Conjugates of naphthalene and dipeptides produce molecular hydrogelators with high efficiency of hydrogelation and superhelical nanofibers. J. Mater. Chem. 17, 850-854 (2007).

[0186] 23. Chen, L., Revel, S., Morris, K., C. Serpell, L. & Adams, D. J. Effect of Molecular Structure on the Properties of Naphthalene-Dipeptide Hydrogelators. Langmuir 26, 13466-13471 (2010).

[0187] 24. DeGrado, W. F. & Lear, J. D. Induction of peptide conformation at apolar water interfaces. 1. A study with model peptides of defined hydrophobic periodicity. J. Am. Chem. Soc. 107, 7684-7689 (1985).

[0188] 25. DeGrado, W. F. Design of Peptides and Proteins. Adv. Protein Chem. 39, 51-124 (1988).

[0189] 26. McCullagh, M., Prytkova, T., Tonzani, S., Winter, N. D. & Schatz, G. C. Modeling Self-Assembly Processes Driven by Nonbonded Interactions in Soft Materials†. J. Phys. Chem. B 112, 10388-10398 (2008).

[0190] 27. Lee, O.-S., Cho, V. & Schatz, G. C. Modeling the Self-Assembly of Peptide Amphiphiles into Fibers Using Coarse-Grained Molecular Dynamics. Nano Lett. 12, 4907-4913 (2012).

[0191] 28. Frederix, P. W. J. M., Ulijn, R. V., Hunt, N. T. & Tuttle, T. Virtual Screening for Dipeptide Aggregation: Toward Predictive Tools for Peptide Self-Assembly. J. Phys. Chem. Lett. 2, 2380-2384 (2011).

[0192] 29. Wu, C., Lei, H. & Duan, Y. Formation of Partially Ordered Oligomers of Amyloidogenic Hexapeptide (NFGAIL) in Aqueous Solution Observed in Molecular Dynamics Simulations. Biophys. J. 87, 3000-3009 (2004).

[0193] 30. Guo, C., Luo, Y., Zhou, R. & Wei, G. Probing the Self-Assembly Mechanism of Diphenylalanine-Based Peptide Nanovesicles and Nanotubes. ACS Nano 6, 3907-3918 (2012).

[0194] 31. 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). doi:10.1039/c3nr02505e

[0195] 32. Thirumalai, D., Klimov, D. & Dima, R. Emerging ideas on the molecular basis of protein and peptide aggregation. Curr. Opin. Struct. Biol. 13, 146-159 (2003).

[0196] 33. Monticelli, L. et al. The MARTINI Coarse-Grained Force Field: Extension to Proteins. J. Chem. Theory Comput. 4, 819-834 (2008).

[0197] 34. Singh, G. & Tieleman, D. P. Using the Wimley-White Hydrophobicity Scale as a Direct Quantitative Test of Force Fields: The MARTINI Coarse-Grained Model. J. Chem. Theory Comput. 7, 2316-2324 (2011).

[0198] 35. De Jong, D. H. et al. Improved Parameters for the Martini Coarse-Grained Protein Force Field. J. Chem. Theory Comput. 9, 687-697 (2013).

[0199] 36. Zaccai, N. R. et al. A de novo peptide hexamer with a mutable channel. Nat. Chem. Biol. 7, 935-941 (2011).

[0200] 37. White, S. H. & Wimley, W. C. Hydrophobic interactions of peptides with membrane interfaces. Biochim. Biophys. Acta BBA—Rev. Biomembr. 1376, 339-352 (1998).

[0201] 38. Wimley, W. C., Creamer, T. P. & White, S. H. Solvation Energies of Amino Acid Side Chains and Backbone in a Family of Host-Guest Pentapeptides†. Biochemistry (Mosc.) 35, 5109-5124 (1996).

[0202] 39. Ash, W. L., Zlomislic, M. R., Oloo, E. O. & Tieleman, D. P. Computer simulations of membrane proteins. Biochim. Biophys. Acta BBA—Biomembr. 1666, 158-189 (2004).

[0203] 40. Yesylevskyy, S. O., Schäfer, L. V., Sengupta, D. & Marrink, S. J. Polarizable Water Model for the Coarse-Grained MARTINI Force Field. PLoS Comput Biel 6, e1000810 (2010).

[0204] 41. Lyon, R. P. & Atkins, W. M. Self-Assembly and Gelation of Oxidized Glutathione in Organic Solvents. J. Am. Chem. Soc. 123, 4408-4413 (2001).

[0205] 42. Han, T. H. et al. Bionanosphere Lithography via Hierarchical Peptide Self-Assembly of Aromatic Triphenylalanine. Small 6, 945-951 (2010).

[0206] 43. Cohen, Y., Avram, L. & Frish, L. Diffusion NMR Spectroscopy in Supramolecular and Combinatorial Chemistry: An Old Parameter—New Insights. Angew. Chem. Int. Ed. 44, 520-554 (2005).

[0207] 44. Pouget, E. et al. Elucidation of the Self-Assembly Pathway of Lanreotide Octapeptide into beta-Sheet Nanotubes: Role of Two Stable Intermediates. J. Am. Chem. Soc. 132, 4230-4241 (2010).

[0208] 45. Barth, A. & Zscherp, C. What Vibrations Tell About Proteins. Q. Rev. Biophys. 35, 369-430 (2002).

[0209] 46. Fleming, S. et al. Assessing the Utility of Infrared Spectroscopy as a Structural Diagnostic Tool for β-Sheets in Self-Assembling Aromatic Peptide Amphiphiles. Langmuir 29, 9510-9515 (2013).

[0210] 47. Fuhrmans, M. & Marrink, S.-J. A tool for the morphological analysis of mixtures of lipids and water in computer simulations. J. Mol. Model. 17, 1755-1766 (2011).

[0211] 48. Georgoulia, P. S. & Glykos, N. M. On the Foldability of Tryptophan-Containing Tetra- and Pentapeptides: An Exhaustive Molecular Dynamics Study. J. Phys. Chem. B 117, 5522-5532 (2013).

[0212] 49. Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 14, 33-38 (1996).

Further References

[0213] (1*) Humphrey, W.; Dalke, A.; Schulten, K. J. MoL Graph. 1996, 14, 33.

[0214] (2*) martinize.py. v2.0, accessed 21 Mar. 2013 on www.cgmartini.nl

[0215] (3*) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E. J. Chem. Theory Comput. 2008, 4, 435.

[0216] (4*) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. J. Chem. Phys. 1984, 81, 3684.

[0217] (5*) Hess, B. J. Chem. Theory Comput. 2008, 4, 116.

[0218] (6*) Marrink, S. J.; Risselada, H. J.; Yefimov, S.; Tieleman, D. P.; de Vries, A. H. J. Phys. Chem. B 2007, 111, 7812.

[0219] (7*) Marrink, S. J.; de Vries, A. H.; Mark, A. E. J. Phys. Chem. B 2004, 108, 750.

[0220] (8*) Yesylevskyy, S. O.; Schäfer, L. V.; Sengupta, D.; Marrink, S. J. PLoS Comput Biol 2010, 6, e1000810.

[0221] (9*) Barth, A.; Zscherp, C. Q. Rev. Biophys. 2002, 35, 369.

[0222] (10*) Fleming, S.; Frederix, P. W. J. M.; Ramos Sasselli, I.; Hunt, N. T.; Ulijn, R. V.; Tuttle, T. Langmuir 2013, 29, 9510.