Linköping Studies in Science and Technology Thesis No. 1254
De Novo
Design and Characterization of Surface Binding
Peptides – Steps toward Creation of Functional Surfaces
Patrik Nygren
LIU-TEK-LIC-2006:35
Division of Molecular Biotechnology Department of Physics, Chemistry and Biology Linköpings universitet, SE-581 83 Linköping, Sweden
© 2006 Patrik Nygren LIU-TEK-LIC-2006:35
ISBN: 91-85523-57-7 ISSN: 0280-7971
The only way - Is all the way
OH - MY – GOD
It's judgement day and I'm not prepared
Everybody out there's running scared
So - Take a little bit off the top
I don't care, just make it stop
Abstract
The ability to create surfaces with well-defined chemical properties is a major research field. One possibility to do this is to design peptides that bind with a specific secondary structure to silica nanoparticles. The peptides discussed in this thesis are constructed to be random coil in solution, but are “forced” to become helical when adsorbed to the particles. The positively charged side-chains on the peptides strongly disfavor an ordered structure in solution due to electrostatic repulsion. When the peptides are introduced to the particles these charges will strongly favor the structure because of ion pair bonding between the peptide and the negatively charged nanoparticles. The peptide-nanoparticle system has been thoroughly investigated by systematic variations of the side-chains. In order to determine which factors that contributes to the induced structure, several peptides with different amino acid sequences have been synthesized. Factors that have been investigated include 1) the positive charge density, 2) distribution of positive charges, 3) negative charge density, 4) increasing hydrophobicity, 5) peptide length, and 6) by incorporating amino acids with different helix propensities. Moreover, pH dependence and the effect of different nanoparticle curvature have also been investigated. It will also be shown that the system can be modified to incorporate a catalytic site that is only active when the helix is formed. This research will increase our understanding of peptide-surface interactions and might be of importance for both nanotechnology and medicine.
Publications
1 Induction of Structure and Chemical Functionality in a de novo
Designed Peptide upon Adsorption to a Silica Nanoparticle. Martin Lundqvist, Patrik Nygren, Bengt-Harald Jonsson and Klas Broo
Submitted to Angewandte Chemie International Edition
2 Optimizing Surface Induced Helical Structure by Rational Changes in
Amino Acid Sequence.
Patrik Nygren, Martin Lundqvist, Klas Broo and Bengt-Harald Jonsson
Abbreviations
AUC Analytical Ultracentrifugation
Boc t-Butoxycarbonyl
CD Circular Dichroism
DCM Dichloromethane
DIPEA N,N’-Diisopropyl Ethylamine
DMF N,N’-Dimethylformamide
EDT Ethanedithiol
Fmoc Fluorenylmethoxycarbonyl
HF Hydrofluoric Acid
HPLC High Performance Liquid Chromatography
MALDI-Tof Matrix-Assisted Laser Desorption Ionization Time of Flight
MeOH Methanol
NMR Nuclear Magnetic Resonance
SPPS Solid Phase Peptide Synthesis
TBTU O-Benzotriazol-1-yl-N,N,N’,N’-tetramethyluronium Tetrafluoroborate
TFA Trifluoroacetic Acid
TIS Triisopropylsilane
UV Ultraviolet
Table of Contents
1 Introduction _______________________________________________________ 1 2 Designig Surfaces Binding Peptides ____________________________________ 3
2.1 Introduction to the Field of Peptides_____________________________________ 3 2.2 The Original Designs _________________________________________________ 6
2.2.1 The initial design, R2A _____________________________________________________ 6 2.2.2 Characterization of R2A ____________________________________________________ 7 2.2.3 Decreasing Charge Density, R1A _____________________________________________ 9 2.2.4 Characterization of R1A ____________________________________________________ 9
2.3 Design Optimization _________________________________________________ 10
2.3.1 Reduced Complexity in R2Ae _______________________________________________ 10 2.3.2 Characterization of R2Ae __________________________________________________ 11 3 Investigating Design Elements _______________________________________ 14
3.1 The Surface Binding Amino Acids _____________________________________ 14
3.1.1 Decreasing the Positive Charge Density _______________________________________ 14 3.1.2 Characterization of R1Ae and R2Z ___________________________________________ 15 3.1.3 Increasing and altering the positive charge _____________________________________ 17 3.1.4 Characterization of R3A and K2A____________________________________________ 18
3.2 The Impact of the Negative Side _______________________________________ 20
3.2.1 Introduction of a Catalytic Site ______________________________________________ 20 3.2.2 Total changes in the negative side ____________________________________________ 24 3.2.3 Characterization of R2AeK and R2AeDN______________________________________ 25
3.3 Increasing Hydrophobicity____________________________________________ 27
3.3.1 Characterization of R2AeL and R2AeV _______________________________________ 27
3.4 Peptide length dependence ____________________________________________ 29
3.4.1 Characterization of R2Ae56 and its parts ______________________________________ 30 4 Methods _________________________________________________________ 32
4.1 Peptide Synthesis____________________________________________________ 32 4.2 Circular Dichroism __________________________________________________ 34 4.3 Analytical Ultracentrifugation_________________________________________ 36 4.4 Silica Nanoparticles__________________________________________________ 37
5 Conclusion and future perspectives ___________________________________ 39 6 Appendix_________________________________________________________ 40 7 References _______________________________________________________ 48 8 Acknowledgements_________________________________________________ 51
1 Introduction
The ability to tailor make surfaces, down to the molecular level, has become a major research field in recent years. The use of well-defined surfaces makes it possible to develop systems on the micro- and nano-scale. With the techniques available this systems can be used to investigate protein-surface interactions [1-4], to create nano-devices such as nanobelts and semi-conductors[5, 6], to design micro-patterns[7], the use of nanopens[8] and nanotrains[9]. It also enables for the understanding of biological systems, such as the growth of organisms on marine surfaces[10-12], the growth of inorganic crystals on SAMs[13] or to understand the growth of ice[14, 15]. It is likely that several applications will have an impact on everyday life.There are several kinds of surfaces that can be used, usually it is pure metal surfaces[16, 17] but also metal oxides[18], plastic[19] and glass[20] should be mentioned. In this project, the surface has been silicon oxide.
By combining organic synthesis and surface science one gets a very powerful tool, and by further including molecular biology, inorganic chemistry, medicine and/or biology the possibilities become almost endless. There are many methods that can be used to analyze chemically modified surfaces, such as IR-spectroscopy[21], x-ray photoelectron spectroscopy (XPS)[22], surface plasmon resonance (SPR)[23], ellipsometry[24] and atomic force microscopy (AFM)[25]. Information that can be provided by these methods includes, thickness of the modification, the interactions with surfactants, the physical properties of the molecules, hydrogen bonding between molecules, charged interactions and structural information.
By creating well-defined surfaces on a molecular scale, one can further introduce functional groups with known distances. This enables for the construction of molecular probes[26], catalysts[27] or selective molecular traps[28]. The achievement of a surface bound or surface initiated system would be useful in large scale chemical synthesis, were the reactant can pass over the surface, react and thereby reduce the by-products present in the purification step, or were the addition of the surface initiates a reaction and the
surface is then easily removed. It can also be used in order to create miniaturized analytical tools, which would be useful in medicine and crime laboratories. Another advantage of a surface attached system is both prize and size. It is useful if the surface area can be washed and reused, and if the surface is micro-patterned, the total surface area exceeds the seen area many times over.
One way to create well-defined surfaces is to use nature as a source of inspiration. By combining nature’s fine solutions and organic chemistry, one gets a powerful tool.
The work done in this thesis is mostly concerned with the use of peptides, and one might ask, “Why use peptides and not other organic molecules”? To answer that question one has to consider multiple aspects. From the organic chemists point of view the first would be that the synthesis of peptides is relatively easy; it can be automatized with a peptide synthesizer machine using a solid support as a starting point and gives a high yield. Secondly, since there is no purification step between the couplings it is less time consuming.
There are other advantages of using peptides for surface modification besides those described above. The natural occurring amino acids contain a number of chemically modifiable side-chains that gives the possibility to incorporate other functionalities. One can also incorporate artificial amino acids and other compounds that can form amide bonds.
Supramolecular systems can be created by designing peptides that binds to each other with high specificity. These systems can pre-organize in solution prior to the surface modification or as will be described in this thesis, adopt a well-defined structure when introduced to the surface.
Investigating peptide folding in the vicinity or at a surface can give vital information on the mechanisms that is working in an organism when protein folding occur and also help in the understanding of how diseases due to missfolding work and how they can be prevented. One might also hypothesize that it can give an insight in how life on earth were initiated, as the most abundant interaction sites in the primordial ooze would have been inorganic surfaces[29].
This is the first time that a peptide has been designed that adopts a specific secondary structure upon binding to a specific surface. This system is also very robust, it is stable at high temperatures and one can introduce reactive sites with only minor changes in helical stability.
2 Designing Surfaces Binding Peptides
2.1 Introduction to the Field of Peptides
Using the amino acids as building blocks for molecular design, produces some limitations but also a vast number of possibilities, as can be seen with naturally occurring enzymes and proteins. Unless one uses non-naturally occurring amino acids, one has 20 different building blocks, which in turn can adopt a secondary structure when connected in a long chain. With the synthetic methods available today those 20 amino acids can, with the use of special protection groups and a cunning chemist, be used to synthesize everything from a protein[30] to a small drug candidate[31]. The possibility to achieve protein folding offers a controlled way to get catalysis[32], attach a molecular probe[33], make self-replication[34] etc.
When designing a peptide one has to consider multiple interactions and different behaviors of the constituting amino acids and although a lot of research has been made in this field[35, 36] it has not been fully examined. A useful tool when trying to predict the outcome of the secondary structure in a peptide chain is to use Chou-Fasmans rules[37] which statistically shows the preference the 20 natural amino acids has for a certain conformation. Although called rules they should be regarded as guidelines.
The most important aspect of a peptide and its behavior is the peptide backbone. A simplified depiction of the backbone is to describe it as composed of amides and α-carbons. The backbone becomes stabilized by the partial double bond of the amides. Interactions between amide carbonyl and amide hydrogen stabilize the different secondary structures encountered in peptides and proteins.
The peptide chain is also affected by the side-chains of the constituting amino acids, their geometry as well as their functional groups. The geometry of side-chains has a tendency to make the amino acids more suitable for different secondary structures and/or interactions with neighboring molecules. The side-chain interactions is the most
important aspect in the folding process of peptides, the formation of salt-bridges, hydrophobic interactions, disulfides and hydrogen bonding all contributes to the stability of secondary and, foremost, tertiary structures.
Hydrophobic interactions may cause a peptide to collapse to form unwanted hydrophobic cores, cysteines can form unwanted disulfides and bind metal ions, and unfortunately, peptides and their side-chains have an attraction towards surfaces. These three attributes affects the functions of peptides and disrupts structures[38, 39].
In this work, peptides have been constructed to use the attraction of surfaces to induce structure, in contradiction to what is generally observed. Utilizing de novo design the peptides are made to behave as unstructured random-coils in solution and when introduced to a negatively charged surface adopt a helical structure (fig. 1). Some of the designs include well placed amino acids that will form an active site[32], only active when the peptides are folded into a helix.
i. ii. iii.
Figure 1. Schematic picture of the random coil to helix peptide. (i.) The peptide in solution. (ii.) Introduction of the negatively charge surface. (iii.) Helix formed on the
surface.
To be able to enhance the understanding of folding and peptide-surface interactions the peptides only contains naturally occurring amino acids (fig. 2), which also have the added benefit of making it possible if needed to use biotechnological methods to create these peptides. The constituting amino acids have been chosen for their chemical and physical properties and their roles in the over-all design have been evaluated by rational changes in the sequence composition.
Picturing the 20 natural amino acids, divided in types of func
tionality; red square contains pos
itive amino acids, green conta
ins
dark blue contains the polar amino acids
and therein light blue
high-lighting cysteine
and the
Given below each are name, abbreviation and pK
a . HO NH 2 O HO NH 2 O HO NH 2 O N NH NH 2 NH NH 2 HN HO NH 2 O HO NH 2 O OH O OH O HO NH 2 O HO NH 2 O SH HO NH 2 O HO NH 2 O HO NH 2 O HO NH 2 O HO NH 2 O OH OH HO NH 2 O NH 2 O HO NH 2 O HO NH 2 O HO NH 2 O HO NH 2 O HO H N O HO NH 2 O HO NH 2 O NH S HO NH 2 O Le u cine Le u, L pK a = -Isol eu cine Ile , I p Ka = -Me th io ni ne M et, M pK a = -Pr o lin e Pr o, P pK a = -P h en yl a lan in e P he, F pK a = -Tr yp to ph an T rp , W p Ka = -Va lin e Va l, V pK a = T yr osin e T yr , Y pK a = 10. 13 Se ri ne Se r, S pK a = ~ 13 Th re on in e T hr , T pK a = ~ 13 Glua m ine Gl n, Q pK a = -A spa ra gi ne A sn, N pK a = -Ala n in e Ala , A pK a = His tid in e His, H pK a = 6 .0 4 A rg in ine A rg, R pK a = 12. 48 L ysin e L ys, K pK a = 1 0.7 9 Gluta m ic aci d Glu, E pK a = 4. 0 7 A spa rt ic ac id A sp , D pK a = 3. 90 Cyste in e Cys, C pK a = 8. 33 Gl yc in e Gl y, G pK a =
-2.2 The Original Designs
2.2.1 The initial design, R2A
The starting point for the design was to choose the length of the peptide. To allow for a obvious helix to form it was decided that the peptide should consist of 28 amino acids, this would under optimal conditions form 7 complete turns[40-42], a helical length often observed in natural proteins.
The peptides were intended to form helical structures upon interaction with a negatively charged silica surface and therefore positively charged amino acids were chosen to favor this interaction. Among the natural amino acids three may have a positive charge;
arginine, histidine, and lysine. The tendency for silica nanoparticles to aggregate at pH below 7 limits the experimental pH range and render the use of positively charged histidine (pKa = 6.04) impossible. Lysine (pKa =10,79) suited the system but the risk of reducing pKa due to a vicinity of other lysines[43] made it a poor candidate as well. Arginine on the other hand seemed as a good choice. It have a high pKa of 12.48, the guanidine group gives it a delocalized charge i.e. a larger charged surface, and low reactivity. To get as much interaction between the peptide and the surface as possible at least one arginine should be present in every helical turn. It was thought that a high concentration of positive charges on one side of the helix would favor the attachment on the surface. Hence, eight arginines were incorporated in the peptide; filling rows 3 and 6 (fig. 3). Besides the favorable interaction with the surface, the eight arginines would prevent the peptide chain from adopting a helical structure in solution due to the electrostatic repulsion that would exist in a helix that was formed in absence of stabilizing negative charges from the silica surface.
Figure 3. Helical wheel depicting R2A. COO -NH3+ Gly7 Leu14 Val21 Leu28 Arg6 Arg13 Arg20 Arg27 Ser5 Gln12 Gln19 Ser26 Thr4 Thr11 Asn18 Ser25 Asp1 Cys8 Asp15 Gln22 Arg24 Arg17 Arg10 Arg3 Leu2 Ile9 Val16 Ala23 Row 1 Row 4 Row 7 Row 3 Row 6 Row 2 Row 5
Hydrophobic amino acids were placed in the adjacent rows (2 and 7, fig. 3). By introducing them next to the arginines and pointing obliquely down towards the surface, they were to fulfill two objectives. Firstly, when the peptide folded on a surface their hydrophobic nature would decrease the interaction between the solution and the arginines and secondly, when helixes had formed they would be able to interact through hydrophobic interactions with other helixes in close vicinity.
When a helix is formed, rows 4 and 5 (fig. 3) will be facing away from the surface towards the solution. Since the solvent would be water based the amino acids incorporated in these two rows would preferably be polar. This enabled them to interact with the solution through hydrogen bonds, and possibly, hydrogen bonds could form between peptides and thereby further favor the helical formation.
Since row 1 (fig. 3) was intended to point away from the surface and face the solution it needed to contain amino acids that would interact well with water and at the same time not be attracted to the negative charges on the silica surface. Silicone oxide at the pH used in this work contains a net negative charge[44] taking this into consideration the most effective amino acids would be either glutamic acid or aspartic acid. Their negatively charged side-chains would be repelled by the surface, this would twist the peptide backbone towards the solution and thereby facilitating the helical formation. In row 1 a cysteine was introduced which could be used for an eventual labeling.
2.2.2 Characterization of R2A
After the synthesis and purification, the peptide was to undergo measurements as to see how successful the design was. The amount of helicity was measured using circular dichroism (CD) in the wavelength region of ~190-260 nm. The measurements were made in room temperature at pH 9.8 and 9 nm particle were used with a ratio of 1:1 (peptide: particle). The spectra were corrected using a reference solution lacking the peptide but otherwise identical.
Figure 4 shows CD-spectra of R2A in solution (black line) which corresponds well with a spectrum of a random-coil structure and agrees with the design. The spectrum of the peptide-particle complex showed a shift indicates the presence of helical structure.
Although not an obvious helix, the shift was enough to investigate the impact that the particles had on the peptides. To determine if the shift was due to a change in the chemical environment or if it was because the peptides had bound to the particles. This
was investigated with analytical ultracentrifugation (AUC) sedimentation equilibrium experiments.
CD-spectrum for R2A
-6 -4 -2 0 2 4 185 192,5 200 207,5 215 222,5 230 237,5 245 252,5 260 Wavelength (nm) D e lt a E p s ilo n
Figure 4. The CD-spectrum of R2A. The black line depicts free peptide in absence of particles and the gray line depicts the peptide particle complex
(9 nm).
The sedimentation equilibrium experiments show that (fig. 5 (i.)) when R2A is in solution it does not sediment until 40000 rpm has been reached, indicating that the peptide is monomeric in solution. The AUC-results (fig. 5(ii.)) show that the peptide is strongly bound to the nanoparticles since it sediments with the particles at 3000 rpm. Thus, the structural change is indeed due to the interaction between the surface and the peptide. (i.) 0,3 0,4 0,5 0,6 0,7 0,8 0,9 6,8 6,85 6,9 6,95 7 7,05 7,1 7,15 Radius Ab so rb an ce (ii.) 0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 6,3 6,35 6,4 6,45 6,5 6,55 6,6 6,65 Radius A b s o rb an ce 3000 3500 3000 4000 3500 6000 4000 8000 6000 12000 8000 20000 12000 40000 20000 44000 40000 46000 48000
2.2.3
Decreasing Charge Density, R1A
In order to increase the helical contentthe results produced by R2A were interpreted as a step in the right direction and therefore used as a template for re-design. The sequence of R1A has many similarities to that of R2A. Rows 2 and 7 containing hydrophobic alanine to shield the surface interacting rows, in rows 4 and 5 polar amino acids, such as tyrosine, glutamine, serine and threonine, were placed to interact with the solution thru hydrogen bonds and row 1were negatively charged.
As for the surface binding rows (3 and
6), there was a major change. While in R2A, both rows had been filled with arginines, R1A only row 3 contained arginine, whereas row 6 was filled with asparagine (fig. 6). The idea was to decrease the net positive charge and thereby reduce the electrostatic repulsion between the positive charged side-chains. The choice of asparagine were made due to the δ-positive nature of the side-chain amide hydrogen which would hopefully be attracted by the surface but not be positive enough to disturb the helical formation due to electrostatic repulsion with the arginines.
The constitution of the hydrophobic rows (2 and 7) differed in one aspect, the complexity of the amino acids were reduced to the use of only one. The peptide R2A (fig. 3) had a mixture of alanine, isoleucine, leucine, and valine in these two rows, which could increase the risk of unwanted hydrophobic interactions. Therefore, in R1A, only alanine that has high helix forming capability was used, keeping a small hydrophobic barrier but reducing hydrophobicity.
Figure 6. Helical wheel depicting R1A COO -NH3+ Gly7 Ala14 Ala21 Ala28 Asn6 Asn13 Asn20 Asn27 Ser5 Gln12 Gln19 Ser26 Thr4 Tyr11 Gln18 Tyr25 Asp1 Cys8 Asp15 Gln22 Arg24 Arg17 Arg10 Arg3 Ala2 Ala9 Ala16 Ala23 Row 4 Row 7 Row 3 Row 6 Row 2 Row 5 Row 1
2.2.4 Characterization of R1A
As for R2A the helicity of R1A were measured with CD-spectroscopy in the wavelength region of ~190-260 nm. This time the ratio between peptide and particle were not 1:1 but
made due to the fact that when the ratio were 1:1 the pH of the sample were altered, even though the particles were dialyzed at the same pH as the sample, this change in pH is probably due to the chemical properties of the particle.
CD-spectrum for R1A
-5 -4 -3 -2 -1 0 1 190 195 199 204 208 213 217 222 226 231 235 240 244 249 253 258 Wavelength (nm) D e lt a E p s ilo n
Figure 7. The CD-spectrum for R1A. The black line depicts free peptide in absence of particles and the gray line depicts the peptide particle complex
(9 nm).
The conformational changes induced by the particles in R1A turned out to be even smaller than for R2A (fig. 7). The reason to why R1A exhibited a lesser helical induction than R2A probably lies in the reduction of positive charges in the peptide, which have decreased the number of possible interaction sites with the charges on the silica surface.
2.3 Design Optimization
2.3.1 Reduced Complexity in R2Ae
As a starting point for the optimization of the design it were presumed that the general idea of the design were correct, meaning that each point in the helical wheel had the right chemical nature and that the problem could be found in the rather complex sequence of the peptide. Therefore, the number of different amino acids used in the design was reduced. Another presumption was that, since R2A had a higher helical content than R1A, it was used as the starting point in the redesign.
Hence, rows 3 and 6 (fig. 8) were filled with arginines as in R2A. For the hydrophobic element, rows 2 and 7, alanine was chosen due to their high helical propensities and rather low hydrophobicity. Glutamine were chosen as the polar element since the side-chain amide group does not contain any acidic hydrogen’s and since it also has a high helical propensity[37], this in comparison with for example serine which has a low helical propensity. COO -NH3+ Cys7 Ala14 Ala21 Ala28 Arg6 Arg13 Arg20 Arg27 Gln5 Gln12 Gln19 Gln26 Gln4 Gln11 Gln18 Gln25 Glu22 Glu15 Glu8 Tyr1 Arg24 Arg17 Arg10 Arg3 Ala2 Ala9 Ala16 Ala23 Ro Row 4 Row 7 Row 3 Row 6 Row 2 Row 5
To decrease the possibility for unwanted interactions and at the same time reduce the number of possible unfavorable interactions the cysteine were removed and row 1 were filled with glutamic acid. Glutamic acid has a higher helical preference than aspartic acid, which is usually found in turns. To facilitate spectroscopic measurements a tyrosine were incorporated in position 1.
w 1
Figure 8. Helical wheel depicting R2Ae.
2.3.2 Characterization of R2Ae
The helical content of R2Ae was measured with CD-spectroscopy at pH 9.0, in the absence of particles and with 9 nm particles. The spectrum of peptide and peptide-particle can be seen in figure 9. The helicity of R2Ae is higher upon addition of nanoparticles than it had been for both R2A and R1A.
Due to the rather obvious helix formed at this pH, corresponding to 27 % helix[45], and with this particle size the peptide were subjected to measurements at both higher, 9.8 and lower, 8.4, pH, with and without particles. This was done to see how well the results corresponds to the measurements done on R1A and R2A but also to test the stability of the formation due to changes in pH and surface area. Percentage of helix for these measurements is shown in figure 10 (the full CD-spectrums are shown in the appendix).
The percentage (fig. 10) indicates that the helix is induced by addition of particle at a wide pH range and that the particle sizes do not influence the formation appreciably. The smaller value for the peptide in absence of particle corresponds well with the design idea. The deviations are not big enough to be interpreted as a helix at either pH.
R2Ae pH 9,0 -5 -4 -3 -2 -1 0 1 200 205 210 215 220 225 230 235 240 245 250 255 260 W a ve length (nm) Del ta E p si lo n
Figure 9. CD-spectrum for R2Ae at pH 9.0. The black line depicts free peptide in absence of particles and the gray line depicts the peptide-particle
complex (9 nm).
Percentage helix of R2Ae
0 5 10 15 20 25 30 35 40 Peptide 6 nm 9 nm 15 nm % h e lix
Figure 10. Percentage helix forR2Ae in absence of particle and with 6, 9 and 15 nm particles. Black = pH 9.8, gray = pH 9.0 and white = 8.4.
It is interesting to note that the highest helical content is observed for the peptide when introduced to the 6 nm particle at the lowest pH (8.4). The results contradict the expectations that the steep curvature of these particles (fig. 11) would put a strain on the backbone hydrogen bonds, which would disfavor a helix.
Figure 11. The length of a fully twisted helix compared with the three different particle sizes.
Encouraged by the good results it was realized that further design optimization was needed. Although there are, numerous possible interactions to take into consideration, in the following chapter the result of five different types of modifications will be presented.
3 Investigating Design Elements
3.1
of arginines were synthesized based on R2Ae, with the reduced complexity of the
The Surface Binding Amino Acids
From the initial investigation, it was clear that the number of positive charges and their distribution strongly affected the tendency to form helical structure on silica surfaces. Therefore, several peptides were designed to investigate these factors in detail.
3.1.1 Decreasing the Positive Charge Density
In the two original designs, the sequence incorporated either one or two rows of arginine. Although the peptide containing two rows showed higher helical content and that the first design optimization therefore contain two rows, it were not proven that it were the best choice. The electrostatic repulsion would be decreased with only one row of arginines and this might still prove to be better, therefore a simpler version with one row
Figure 12.To the left, helical wheel depicting R1Ae and to the right R2Z. COO -NH3+ Ala7 Ala14 Ala21 Ala28 Asn6 Asn13 Asn20 Asn27 Gln5 Lys12 Gln19 Gln26 Gln4 Gln11 Gln18 Gln25 Glu1 His8 Glu15 Tyr22 Arg24 Arg17 Arg10 Arg3 Ala2 Ala9 Ala16 Ala23 Row 1 Row 4 Row 7 Row 3 Row 6 Row 2 Row 5 COO -NH3+ Ala7 Ala14 Ala21 Ala28 Asn6 Arg13 Asn20 Arg27 Gln5 Gln12 Gln19 Gln26 Gln4 Gln11Gln18 Gln25 Asn24 Arg17 Asn10 Arg4 Ala3 Ala9 Ala16 Ala23 Glu1 Glu8 Glu15 Glu22 Row 1 Row 4 Row 7 Row 3 Row 6 Row 2 Row 5
sequence but incorporating asparagine in row 6 instead of arginine (fig. 12). Row 3 were filled with arginines, alanines in rows 7 and 3, in rows 4 and 5 glutamine were placed and glutamic acid in row 1. The peptide was named R1Ae, meaning one row of arginines and a simple amino acid composition.
To further investigate the influence that a reduced positive charged peptide would have on the helical formation a peptide, R2Z, were constructed. R2Z have the same amino acid composition in rows 1, 2, 4, 5 and 7 as R2Ae with the reduced complexity, but differs in rows 3 and 7 in the aspect that instead of filled rows of arginine every other amino acid are arginine and every other is asparagine. This makes the charges located in a zigzag pattern (fig. 12).
3.1.2
Characterization of R1Ae and R2Z
Figure 13 shows the CD-spectrums of R1Ae and R2Z in pH 9.0 buffer with and without particles. The addition of particles produce a conformational change in the peptides pushing the minima towards 208 nm and another minimum is forming at 222 nm. The changes in R1Ae are smaller than the changes in R2Z and compared with R2Ae both form less helical structures. R1Ae corresponds to 6 % helix in absence of particle and with particle added, 8 %; none of these values describes an obvious helix[45]. The same values for R2Z are 7 % when the peptide is free in solution and 22 % when the particles have been added.
The peptides were also investigated with all three particle sizes and at pH 9.8 and 8.4 (full CD-spectrums can be found in the appendix). The percentage helix values can be found in figure 14. For most of the systems, the helical formation is higher for R2Z, although it does not produce as good helixes as R2Ae. The values lies in between 8 and 20 percentage, with the exception of the peptide-particle complex formed for R2Z with 9 nm particles in the 9.8 buffer. This might be explained when taking into consideration that the surface charge is largest at this pH. As to why the content for the 15 nm particle is not as high as for the other particle sizes, can be due to the small repulsive force generated between the surface and the carbonyl oxygen in the asparagine. This small repulsive force might also explain why R1Ae exhibits such low structural affect when introduced to the particles.
R1Ae and R2Z pH 9,0 -6 -5 -4 -3 -2 -1 0 1 200 205 210 215 220 225 230 235 240 245 250 255 260 W a ve le ngth (nm) D e lt a E p s ilon
Figure 13. CD-spectra of R1Ae and R2Z in pH 9.0. Black dashed line = R1Ae in absence of particle. Black line = R1Ae with 9 nm particles. Gray dashed line = R2Z in absence of
particles. Gray line = R2Z with 9 nm particles.
Percentage helix of R1Ae and R2Z
0 5 10 15 20 25 30 35 Peptide 6 nm 9 nm 15 nm % h e lix
Figure 14. Percentage helix for R1Ae and R2Z, in absence of particle and with 6, 9 and 15 nm particles. Left black = R1Ae in pH 9.8. Left gray = R1Ae in pH 9.0. Left white = R1Ae in pH 8.4. Right black = R2Z in pH 9.8. Right gray = R2Z in pH 9.0.
3.1.3
Increasing and altering the positive charge
To investigate the contribution the positive charges have on the formation of the helixes further, a peptide with an increased number of arginines were synthesized, incorporating a third row of arginines. This would give yet another handle for the peptide to adsorb to the surface. By introducing a third row the overall design of the peptide had to be changed, as one row that were occupied by alanines, glutamines or the glutamic acids were lost.
It was believed that the repulsion the glutamic acids exhibited from the surface were a crucial part of the design and that the hydrophobic element in the design played a larger role than the hydrophilic. Therefore, the design of R3A, besides having three rows of arginines, was to contain two rows of alanine and two rows of glutamic acid (fig 15).
o far, arginine has been used as the anchor to get the peptide to adsorb to the surface an
Figure 15. To the left, helical wheel depicting R3A and to the right K2A. Glu7 Glu14Glu21 Tyr28 Arg5 Arg12 Arg19 Arg26 Glu4Glu11 Glu18Glu25 Arg6 Arg13 Arg20 Arg27 C-Term. N-Term. Arg2 Arg9 Arg16 Arg23 Ala24 Ala17 Ala10 Ala3 Cys1 Ala8 Ala15 Ala22 Row 4 Row 3 Row 5 Row 6 Row 7 Row 1 Row 2 C-Term. N-Term. Ala7 Ala14 Ala21 Ala28 Lys6 Lys13 Lys20 Lys27 Gln5 Gln12 Gln19 Gln26 Gln4 Gln11 Gln18 Gln25 Glu1 Glu8 Glu15 Glu22 Lys24 Lys17 Lys10 Lys3 Ala2 Ala9 Ala16 Ala23 Ro Row 4 Row 7 Row 3 Row 6 Row 2 Row 5 w 1 S
d this has proven to be successful. In the pH range used, lysine, with a pKa of 10.79, would also be positively charged and by replacing the arginines with lysine, the impact that the guanidine group have on the structural change would be tested. The design of K2A were a duplicate of R2Ae with the exception of the replaced arginines (fig. 15)
3.1.4
Characterization of R3A and K2A
CD-spectrum of R3A (fig. 16) in pH 9.0 corresponds to a well-defined helix, with a calculated helix content of 73 % for the peptide interacting with the particle while free R3A in solution only had 4 %[45]. This indicated that more arginines did not produce a too high repulsion but instead increased the binding to the particles and also helped in the structural formation. Looking at how R3A behaves with changes in pH and surface size it is apparent that a smaller particle induces more structure in the peptide and that a pH of 9.0 induces the most apparent helical structure at all particle sizes (fig. 17). This might be because as the particle size increases the number possible attachment points for the
arginine increases and thereby increasing the strain in the backbone, which would distort the helix. As to why pH 9.0 gives the best helical content most probably has to do with the equilibrium between the charges on the arginines and on the surface. As the pH decreases, the surface charge becomes reduced thus decreasing the number of possible interaction sites, which disfavors formation, and when pH increases, the surface becomes more negatively charge that enables the arginines to interact at a larger interaction surface, which smears the peptide to the surface, rather than induce structure.
R3A and K2A pH 9,0
-8 -7 -6 -5 -4 -3 -2 -1 0 1 200 205 210 215 220 225 230 235 240 245 250 255 260 W a ve le ngth (nm) D e lt a E p s ilon
Figure 16. CD-spectra of R3A and K2A in pH 9.0. Black dashed line = R3A in absence of particle. Black line = R3A with 9 nm particles. Gray dashed line = K2A in
K2A when introduced to the silica particles clearly changes toward a helical structure minima are forming at both 208 and 222 nm (fig. 16), although the overall change is almost not visible, the calculate helical content goes from 8 % when K2A acts as a free peptide to 12 % when particles were added. Looking at the CD-measurements for K2A at the three pHs used it is apparent that it gets a higher helical content as pH is increases (fig. 17).
Percentage helix of R3A and K2A
0 10 20 30 40 50 60 70 80 90 Peptide 6 nm 9 nm 15 nm % he lix
Figure 17. Percentage helix for R3A and K2A, in absence of particle and with 6, 9
and 15 nm particles. Left black = R3A in pH 9.81. Left gray = R3A in pH 9.0. Left
white = R3A in pH 8.4. Right black = K2A in pH 9.8. Right gray = K2A in pH 9.0. Right white = K2A in pH 8.4.
The helical formation of K2A also increases with decreasing particle sizes. The increased helical structure induced by the higher pH most probably derives from the increased negative charge of the surface making it easier for the amine to find an attachment point. The reason to why smaller particles give a better helical formation might lie in the fact the lysines in close vicinity tend to reduce the pKa of the side-chain amine[43]. As this happened the side-chains become deprotonated that could lead to a
helical formation due to hydrophobic interactions and therefore the need for a large surface interaction diminishes.
3.2
put in a i, i+4 position in a helix this pair produces a active site for ester hydrolysis[32]
The Impact of the Negative Side
In the above-described designs, glutamic acid and glutamine have been used to interact with the solution. Glutamic acid would also be repelled by the surface that would help inducing the helix formation as it twists the peptide backbone. To test this, four peptides were synthesized containing various changes in these rows.
3.2.1
Introduction of a Catalytic Site
A catalytic His-Lys pair was incorporated in the sequences of R2Ae and R3A. When
where the imidazole of the histidine will initiate the hydrolysis by an attack on the ester carbonyl and the lysine amine would stabilize the negative charge that will develop in the
R2Aecat and R3Acat pH 9,0
-4 -3,5 -3 -2,5 -2 -1,5 -1 -0,5 0 0,5 200 205 210 215 220 225 230 235 240 245 250 255 260 W a ve le ngth (nm) D e lt a E p s ilon
Figure 18. CD-spectra of R2Aecat and R3Acat in pH 9.0. Black dashed line = R2Aecat in absence of particle. Black line = R2Aecat with 9 nm particles. Gray
dashed line = R3Acat in absence of particles. Gray line = R3Acat with 9 nm particles.
acylintermediate. In the peptide based on R2A, called R2Aecat, the histidine was inserted in position 15 and the lysine in position 19. R3Acat, based on R3A, the histidine was inserted in position 14 and the lysine in position 18. The His-Lys pair would also disrupt the negative side of the peptides
It is evident that the introduction of the His-Lys pair does not affect the helical content upon addition of particles, compared to the peptides without the catalytic pair. R2Aecat corresponds to 7 % helix as a free peptide and 20 % helix[45, 46], when particles are added. The same calculations for R3Acat give 7% helix as peptide in absence of particle and with particle 27 % helix.
Percentage helix of R2Aecat and R3Acat
0 5 10 15 20 25 30 35 40 Peptide 6 nm 9 nm 15 nm % he lix
Figure 19. Percentage helix for R2Aecat and R3Acat, in absence of particle and with 6 9
D-measurements with varied pH and particle sizes showed that both peptides exhibits a great deal of helical content in the whole range and that the percentage values lies in the
,
and 15 nm particles. Left black = R2Aecat in pH 9.82. Left gray = R2Aecat in pH 9.0.
Left white = R2Aecat in pH 8.4. Right black = R3Acat in pH 9.8. Right gray = R3Acat in
pH 9.0. Right white = R3Acat in pH 8.42.
sa
showed that th
me range independent of pH and particle size, with the exception of R2Aecat that have a obvious increase when subjected to the 15 nm particles at the higher pHs.
R2Aecat were further analyzed with AUC as to see whether it was bound to the particles or not. The AUC measurement of the peptide in absence of particles
e peptide were in solution, whereas the measurement of the peptide-particle system showed that the peptide have adsorbed to the colloidal particles (fig. 20).
(i.) 0,185 0,19 (ii.) 0,4 -0,1 0 0,1 0,2 0,3 6,3 6,35 6,4 6,45 6,5 6,55 6,6 2500 0,155 0,16 0,165 0,17 0,175 0,18 6,8 6,85 6,9 6,95 7 7,05 7,1 7,15 2500 3500 3500 4500 4500 8000 8000
Figure 20. (i.) AUC results of free R2Aecat. (ii.) AUC results of R2Aecat with 9 nm particles.
To further analyze the stability of e-particle system, a 1 ml sample of R2Aecat mixed with 9 nm particles were dialyzed for 60 minutes in 1000 ml, pH 9,0, c
the peptid
arbonate buffer at room temperature. An equal amount of sample was treated with a pH 2 hydrochloric acid solution under the same conditions. After these treatments were both samples analyzed with AUC (fig. 21).
(i.) 0,6 0,7 (ii.) 0 0,05 0,1 0,15 0,2 0,25 6,85 6,9 6,95 7 7,05 7,1 7,15 2500 2500 0 0,1 0,2 0,3 0,4 0,5 6,3 6,35 6,4 6,45 6,5 6,55 6,6 6,65 3500 3500 4500 4500 8000 8000
Figure 21. (i.) AUC results of R2Aecat dialyzed for 1 hour. (ii.) AUC results of R2Aecat after treatment with hydrochloric acid
It is apparent that t hour of dialysis, as
shown in figure 21 (i.) where the peptides sediments at the same rotor speed as the p
reatment displays a free peptide (fig. 21(ii.). Whether this is because the he peptide-particle complex is retained after one
articles.
Treating the complex with strong acid breaks the interaction and the results obtained after this t
peptide-particle interaction has been broken and they exist in the solution, or if the particles have dissolved, have not been investigated.
Figure 22. (4-sulfamoyl-benzoylamino)-acetic acid 3-nitorphenyl ester
Since th or ester
hydrolysis R2Aecat were chosen to investigate if it worked. UV-measurements were m
e His-Lys pair incorporated in the design would work as a reactive site f
ade on the buffer, buffer containing particles, buffer containing R2Aecat and the peptide-particle complex, all containing the substrate (4-sulfamoyl-benzoylamino)-acetic acid 3-nitorphenyl ester (fig. 22).
Figure 23. Ester bond hydrolysis by: the peptide-particle complex in buffer (◊); peptide in buffer (▲); particles in buffer (■); buffer (♦). The reaction product m-nitro phenol was
The peptide-particle complex sh ase over the background reaction. Addition of more substrate showed that the activity were not terminated due to amide fo
rement showed
monitored at A357.6.
owed a 10-fold incre
rmation on the lysine amine which has been observed in similar systems[46, 47]. This could also be verified by MALDI-Tof which did not show any sulfonylated product after the reaction. This finding makes it possible to use the system as a man-made catalyst and if combined with a gentler way of detaching the peptides from the particle it could be used as a molecular switch. These results are further discussed in Paper 1.
The temperature dependence of R2ecat was determined by raising the temperature ten ºC at a time starting at 25 ºC and going to 75 ºC (fig. 24). The CD-measu
Temperature dependence 25 30 0 5 10 15 20 25 35 45 55 65 75 25 Te m pe ra ture (ºC) % h e lix
Figure 24. Influence of temperature on R2Aeca depicted as percentage helix. Black = R2Aecat adsorbed to 9 nm particles Gray = R2Aecat in absence of particle. White =
R2Aecat after the sample had cooled to 25°C again).
A d
again. The measu ation were fully
reversible. This reversibility could make the system useful as a temperature-induced c
R3Acat showed that the system could
si
egative si
fter the temperature increasement the sample were cooled to 25 °C and measure rement revealed that the disruption of the conform
atalyst, if the catalytic activity will diminish as the temperature increases. It would therefore be interesting to investigate the temperature dependence of the activity.
3.2.2 Total changes in the negative side
Since the results given by R2Aecat andsustain smaller changes in the negative de it were interesting to see if it were possible to alter it more.
As a first approach R2AeK were synthesized with an increased number of positive charges located in the n
de compared with the catalytic peptides. In position 8, 11, 12, 15 and 19 lysine was introduced (fig. 25). Lysine was chosen over arginine since
COO -NH3+ Ala1 Ala8 Ala15 Ala22 Arg2 Arg9 Arg16 Arg23 Gln3 Lys10 Lys17 Gln24 Gln4 Gln11 Lys18 Gln25 Glu7 Lys14 Lys21 Glu28 Arg26 Arg19 Arg12 Arg5 Ala6 Ala13 Ala20 Ala27 Row 1 Row 4 Row 7 Row 3 Row 6 Row 2 Row 5
measurement on K2A showed that the side-chain did not have a large effect on the structural formation. In the case of a pKa decrease it would most probably favour the formation. If this major change in the peptide would be able to sustain the helical formation when introduced to the particles it could also be used as a decarboxylation agent for oxaloacetate[43].
In the design of R2Ae most of the amino acids have a propensity for helixes and to see h
3.2.3 Characterization of R2AeK and R2AeDN
structure independent of w
ow a decrease in this propensity would affect the formation, R2AeDN were made. In this peptide row 1 were filled with aspartic acid and asparagine were inserted in rows 4 and 5. This replacement exchanged the glutamic acid and glutamines residues, both having high preferences for helixes whereas aspartic acid and asparagine both are considered to be helix breakers[37].
Introduction of particles to R2AeK showed no changes in the
hich pH or which particle sizes were used (fig. 26 and 27). The probable cause for this is the lysines which initially will be attracted by the surface making it impossible to form
Figure 26. CD-spectra of R2AeK and R2AeDN in pH 9.0. Black dashed line = R2AeK and R2AeDN pH 9,0
-5 -4,5 -4 -3,5 -3 -2,5 -2 -1,5 -1 -0,5 0 0,5 200 205 210 215 220 225 230 235 240 245 250 255 26 W a ve le ngth (nm) D e lt a E p s ilon 0
a helix. This means that both the arginines and the lysines might be attached to the surface, which would make the peptide smeared to the particle with both the positive and the former the negative side. This in turn would make the helical formation impossible.
R2AeDN on the other hand displays a rather fine helical formation upon addition of the particles in pH 9.8. The degree of helical formation lies in the same vicinity as R2Aecat and R3Acat, were R2Aecat shown good enough helix formation to enable a catalytic activity. As for most of the other peptides investigated R2AeDN exhibits a lower degree of helicity when introduced to the particles at lower pH. At pH 9.8 the percentage of helix does not change to a higher extent over the different particle sizes. Compared with the measurements on both 8.4 and 9.0 were helix formation is almost nonexistent, with the exception of pH 9 when 9 nm particles is added.
Percentage helix of R2AeK and R2AeDN
0 5 10 15 20 25 30 Peptide 6 nm 9 nm 15 nm % h e li x
Figure 27. Percentage helix for R2AeK and R2AeDN, in absence of particle and with 6, 9 and 15 nm particles. Left black = R2AeK in pH 9.8. Left gray = R2AeK in pH 9.0. Left white = R2AeK in pH 8.4. Right black = R2AeDN in pH 9.8. Right gray = R2AeDN in pH
9.0. Right white = R2AeDN in pH 8.43
3.3 Increasing Hydrophobicity
Alanine have been used in the earlier designs since it has a high helical propensity and a low hydrophobicity, promoting helical formation and decreasing the risk of getting the peptides to interact through hydrophobic interactions. As with all other elements, the role of the alanines was to be investigated and this was done by synthesizing two peptides with a higher degree of hydrophobicity. The first were synthesized exchanging the alanines with leucine. Leucine has, as alanine, a high preference for the helix structure and the exchange would therefore test only the influence of the hydrophobicity. The peptide, based of R2Ae, called R2AeL where L denotes the exchange of alanine to leucine. A second peptide, R2AeV, was synthesized with valine placed in the positions otherwise held by alanine. Valine has a higher preference for the β-sheet structure[37] compared to helical structures, but is also more hydrophobic than alanine.
3.3.1 Characterization of R2AeL and R2AeV
Mixing R2AeL with 9 nm particles give the CD-spectra of an obvious helix (fig. 28),
R2AeL and R2AeV pH 9,0
-5 -4,5 -4 -3,5 -3 -2,5 -2 -1,5 -1 -0,5 0 0,5 200 205 210 215 220 225 230 235 240 245 250 255 26 W a ve le ngth (nm) D e lt a E p s ilon 0
Figure 28. CD-spectra of R2AeL and R2AeV in pH 9.0. Black dashed line = R2AeL in absence of particle. Black line = R2AeL with 9 nm particles. Gray
compared with the template, R2Ae; it is even better. Analysis of R2AeL at all pH and with all particles shows that it has good helical content in the whole range. At pH 9.0 it is has a percentage of helix above 25 with all particle sizes. At pH 8.4, a somewhat lesser degree of helicity is observed but obvious helixes can bee seen independent of particle size.
The most promising result is given for R2AeL in pH 9.8 were the percentage helix lies between 89 and 96, which corresponds to an almost perfect helixes[45]. This shows that the hydrophobic interactions facilitates the formation of the helix and probably makes the peptides closely packed on the surface of the particles.
Percentage helix of R2AeL and R2AeV
0 20 40 60 80 100 120 Peptide 6 nm 9 nm 15 nm % h e li x
Figure 29. Percentage helix for R2AeL and R2AeV, in absence of particle and with 6, 9 and 15 nm particles. Left black = R2AeL in pH 9.8. Left gray = R2AeL in pH 9.0. Left white = R2AeL in pH 8.4. Right black = R2AeV in pH 9.8. Right gray = R2AeV in pH
9.0. Right white = R2AeV in pH 8.4
The valine containing peptide shows a small conformational change at pH 9.8. At pH 9.0 it has a higher helical content as a peptide in absence of particle than when particles are added, and is unaffected by the introduction of particles at pH 8.4. Most probably the chemical nature of valine has a too great influence on the entire chain and therefore the peptide is almost unaffected by the addition of particles. At pH 9.8, the increased surface
charge becomes large enough to compete with the valine making the peptide fold into a helix, although increase in structure is almost negligible.
By combining the results gained from R2AeL and R2AeV it can be said that increased hydrophobicity affects the formation in a positive way but that the amino acids propensity for helix structure plays an important role.
3.4 Peptide length dependence
Variations in the length of a polypeptide chain have a tendency to influence the stability of a formed helix[40-42], where longer length usually means higher stability. As to see how different length would affect the system, a peptide twice as long as the template were synthesized, R2Ae56. This peptide were to contain exactly the same amino acid composition as the template, but in order to facilitate the SPPS the peptide were divided in two parts synthesized separately.
The two fragments, A and B, were after synthesis going to be coupled using a Kent ligation[30]. Part A were synthesized as the first 25 amino acids in the sequence using an orthogonally protected glutamic acid for the first coupling, this would yield a glutamine when cleaved from the resin. The orthogonal protective group was removed and the carboxyl was activated with benzylmercaptan before cleavage. The amino terminal on part B, containing the remaining amino acids, was left as a free amine. These two changes from the normal synthetic rout make a Kent ligation possible (fig. 30).
Part A OH O Part A S O Ph HS Ph -S Ph Part A S O Ph Part A S O Part B NH2 H2N Part B HS Part A N H O Part B SH
3.4.1
Characterization of R2Ae56 and its parts
Although the interaction between a longer peptide and the particles were the goal of this investigation, the two parts used to synthesis it was also investigated as to how they behaved when introduced to the particles.
CD-measurements revealed that R2Ae56 and both part A, after addition of benzylmercaptan, and B formed well-defined helixes when introduced to the particles. R2Ae56 showed a smaller shift in the helical region than expected due to its length, this might be due just to the length and the influence the curvature of the particles will have on the helix. The curvature will put a strain on the helical backbone that will deform the helix. Even though this might have happened, R2Ae56 still showed a higher helical content than the template R2Ae.
R2Ae56 and part A and B pH 9,0
-8 -7 -6 -5 -4 -3 -2 -1 0 1 200 205 210 215 220 225 230 235 240 245 250 255 260 W a ve le ngth (nm) D e lt a E p s ilon
Figure 31. CD-spectra of R2Ae56 and part A and B in pH 9.0. Black dashed line = R2Ae56 in absence of particle. Black line = R2Ae56 with 9 nm particles. Dark gray dashed line = part A in absence of particles. Dark gray line = part A with 9 nm particles.
Light gray dashed line = part B in absence of particles. Light gray line = part B with 9 nm particles.
Part A shows more helicity than both R2Ae and R2Ae56, although it is shorter. This might have something to do with the benzyl group attached to the C-terminal as it might work as a helix inducer or a helix C-capping group[36]. As for most other peptides
described part A exhibits higher helical contents at the higher pH, but its noteworthy that it has as good helical content as R2Ae in pH 8.4.
For part B the high helical content might be due to the extended length factor, as it is composed of 31 amino acids compared with 28 as the earlier described peptides has been, and that this length still is short enough to not be to influenced by the curvature. The helical content for part B is high throughout the pHs, although it is clear that pH 8.4 have a lesser affect on the formation than the higher pHs. It should be noted that part B have a free amine in the N-terminal, and even though this is present, the helical formation is not disturbed. The reason to this is probably the dipole moment produced by the helix that will decrease the pKa of the amine making it deprotonated and therefore unable to interact with the surface.
Percentage helix of R2Ae56 and part A and B
0 10 20 30 40 50 60 70 Peptide 6 nm 9 nm 15 nm % h e li x
Figure 32. Percentage helix for R2Ae56 and part A and B, in absence of particle and with 6, 9 and 15 nm particles. Left black is omitted due to lack of material. Left gray = R2Ae56 in pH 9.0. Left white = R2Ae56 in pH 8.4. Middle black = part A in pH 9.8. Middle gray = part A in pH 9.0. Middle left = part A in pH 8.4. Right black = part B in
4 Methods
4.1 Peptide Synthesis
All peptides investigated in this work have been synthesized using solid-phase peptide synthesis (SPPS). The groundbreaking work of solid-phase synthesis was developed by Merrifield[48, 49] et al. and is now used in the synthesis of biopolymers, combinatorial solid-phase organic chemistry, synthesis of natural products, catalyst selection, chemical ligation and material development[50]. The solid supports are usually polystyrene beads with a low degree of cross-linking by divinylbenzene, incorporating a linker with one of many different functional groups used as an anchor for the synthesis. The low cross-linking used enables the resins to swell which in turn allows the growing peptide chain to accommodate therein. There are many advantages of solid phase synthesis over the solution based. 1) high excess of added reagents drives reactions to completion, 2) easy removal of surplus as well as reaction by-products by washing, 3) negligible loss of product during synthesis since it is bound to the resin, 4) use of different protection groups enables for branched peptides and a wide variety of functionalization, 5) the possibility to make the synthesis automated and thereby decrease manual work.
The increasing interest in peptide chemistry and the use of SPPS forced the development of protection groups to move from the unpleasant Boc-chemistry, which utilizes TFA and HF for deprotection respectively cleavage, to the more user friendly Fmoc-chemistry, where piperidine is used for deprotection and TFA is used to cleave the peptide from the resin. Since most peptides are synthesized as long chains of unbranched amide bonds, the use of automation for synthesis is preferred. In this work, peptides have been synthesized on a PioneerTM Peptide Synthesis System, which is fully automated. This SPPS-system uses DMF as wash and base-solvent, Fmoc deprotection have been done with 20% piperidine. Activators for coupling are TBTU and DIPEA.
OH O H C R1 N H N H H C R2 O O O H C R1 N H X H2N H C R2 Y O O O H C R1 N H X Fmoc N H H C R2 Y O Fmoc N H H C R2 Y O Act O O H C R1 N H Fmoc X Fmoc N H H C R2 Y O OH O O H C R1 N H Fmoc X Activator Piperidine Activation Deprotection Coupling Repeat of deprotection and coupling cycles
Final deprotection Cleavage O H C Rn H2N
Figure 33. Schematic picture of SPPS. R1, R2 and Rn represent the amino acid
side-chains. X, Y and Z represent eventual protection groups.
During the synthesis, amino acids are kept in test tubes placed in an amino acid dispenser sledge. Before coupling to the resin, the amino acids are pre-activated with TBTU and DIPEA. Thereafter the activated amino acids are flown past the resin several times during the coupling cycle. This cycle is followed by a washing cycle to rinse the resin from unreacted amino acids and the reaction by-products. After wash, there is an Fmoc deprotection cycle preparing the growing chain to couple the next amino acid. Unless the newly added amino acid is the last of the sequence in which case one of the
the chain, 2) Fmoc off, giving a free amine at the terminal, or 3) cap on, where the N-terminal amine is reacted with acetic acid anhydride giving an acetoamide N-N-terminal. (Fig. 33)
Before cleavage, the resin is removed from the synthesizer system and washed with acetic acid, DCM and MeOH, and dried under high vacuum. Depending on the constituting amino acids, the cleavage cocktail used differs but the main part is TFA. To prevent that the side-chain protection groups undergoes unwanted side reactions with the peptide chain, scavengers such as TIS, water and in case the sequence contains cysteine EDT is added to the cocktail (fig 34). The resin is removed by filtration and washed with TFA. To ease the precipitation the filtrate TFA is evaporated with N2-gas before cold ether is added. The precipitate is centrifuged and the ether phase is removed, the peptide is dissolved in water and lyophilized before purification.
PEG CH2N H O (CH2)4 O OMe OMe N H O H N O O O N H O H N O O O H TFA TFA TFA Piperidine
Figure 34. The Fmoc-PAL-PEG-PS resin and showing are also the cleavage sites for piperidine and TFA.
4.2 Circular Dichroism
When one wants to deduce the secondary structure of a peptide there are three different techniques that precedes all others; X-ray crystallography, nuclear magnetic resonance (NMR) and circular dichroism (CD). Of these, X-ray crystallography gives the best results for proteins that can be crystallized, that is not possible with the system investigated in this thesis. NMR requires high sample concentration; it is time consuming and sometimes requires isotope labeled nuclei’s (13C, 15N) which is very expensive. CD measurements are rapid and require small amounts of analyst, which may be recovered for later use. CD measures the difference in absorption of left and right circularly polarized light from plane-polarized light. Optically active chromophores make this
absorption effect occur when 1) its structure permits it, or 2) if it is covalently coupled to a chiral center, or 3) it is placed in an asymmetric environment.[51]
180 190 200 210 220 230 240 250 W a ve le ngth (nm ) De lt a E p si lo n α-helix β-sheet Random coil
Figure 35. Depicting the typical CD-spectrum of the different structures in the far-UV region.
The spectral range most often used in investigation of peptides and proteins lies between 180-300 nm. The region below 240 nm is referred to as far-UV and the region from 260-300 nm is referred to as near-UV. In the far-UV region, information is gained of the secondary structure of the peptide chain. The different spectra’s obtained in this region derives from the orientation of the back-bone amide in the different secondary structures (i.e. α-helix, β-sheet and turn, figure 35) CD makes use of the fact that left and right polarized light exhibits chirality and by doing that interacts differently with chiral molecules. The dominating absorption in a CD spectrum in the range of 190-240 nm derives from the amide bond in the peptide backbone. This adsorption is different for the different secondary structures.
The values of CD spectroscopy are often given in ellipticity. The ellipticity is calculated using this formula:
where Δε is the difference in molar extinction coefficients of right, εR, and left, εL, polarized light (Δε = εR – εL).
A right-handed α-helix has a distinctive double minimum at 208 and 222 nm. The minimum at 222 nm is a good measurement of the helical content of a peptide that contains mostly of helices.
The CD data gathered in this thesis were collected at 0.5 nm intervals with an integration time of 2 s for the region between ~190-240 nm and with 2 nm intervals between 240-260 nm. The spectrum represents an average of three consecutive scans and before summation; the three separate scans were compared to detect possible alterations of the sample during the scan period.
4.3 Analytical Ultracentrifugation
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 6 6,2 6,4 6,6 6,8 7 Ra dius (cm ) A b so rb an ce 2000 15000 50000Figure 36. Schematic picture of concentration gradients at three different rotor speeds.
The simplest way of describing the principles of analytical ultracentrifugation (AUC) is to compare it with when sand is dropped into a glass of water. The sand will descend at different speeds down to the bottom of the glass depending on the relative size and weight. The same is true for macromolecules and particles although in their cases, an external force will be needed, and in AUC, this force is the centrifugal. Varying the
conditions used in the AUC makes it possible to determine sample purity, molecular weight, analysis of associating systems, detection of conformational changes but usually AUC is used for equilibrium sedimentation and velocity sedimentation. In this work equilibrium sedimentation has been used[52].
The centrifugal force affecting the sample molecules will force them to move towards the outer wall of the sample compartment. At the same time an opposing, diffusion, force will retain the molecules in the solution. After some time the two forces will reach equilibrium, which in turn will reestablish a uniform concentration. By monitoring the changes in absorbance of the sample the sedimentation rate of the sample can be depicted as a function of absorbance versus radius (fig. 36) and this can give information of the above mentioned parameters[53].
In this work rotor speeds between 2500-50000 rpm have been used. Equilibrium was reached after 20 hours and the sedimentation were monitored at 230, 240 and 280 nm. The samples had a 1:1 ratio between peptide and particle. The temperature was kept at 20 °C for all experiments
4.4 Silica Nanoparticles
The silica nanoparticles used as surfaces in these studies were kindly provided by EKA-Chemicals, Stenungsund, Sweden, as colloidal solutions. The particles are stable in a narrow pH range from around 7.5 to 10.5 (fig. 37) and in moderate salt concentrations. The surface charge density of the three particle sizes (6, 9 and 15 nm) used in this thesis is the same at each pH[54]. At pH 8.4, the surface charge is around 0.5 OHӨ/nm-2, at pH 9.0 around 0.7 OHӨ/nm-2, and at pH 9.8 around 1 OHӨ/nm-2[55].
Silica nanoparticle solutions are transparent and the concentrations used in the CD-measurements scatters light negligibly[39, 56, 57], making them suitable not only for CD but also for techniques such as IR and fluorescence.
5 Conclusion and future perspectives
This thesis have shown that it is possible to, by rational design, induce a secondary structure in a polypeptide chain upon binding to an inorganic surface. By controlled changes in the amino acid sequence, the contributions of the constituting amino acids and their role in the conformational change have been verified. It has also been shown, through AUC-measurements and by dialysis, that the peptide adsorbs strongly to the silica particles used to initiate the induced structure. Furthermore, it has been shown that the conformational changes are stable in a wide range of pH as well as through a wide temperature range. As a final test it have been shown that induction of structure in a peptide can be used to get catalytic activity.It seems that among the most important factors contributing to induce a helical structure in a peptide with the use of silica nanoparticles is. A) a large number of arginines placed close together, B) a higher degree of hydrophobicity and C) the use of amino acids with high helical propensities.
Although the system has been rather extensively investigated, there are still many openings for further experiments. For example, one can replace the silicon oxide with gold nanoparticles functionalized with a negatively charged thiol-containing reagent or use a macroscopic surface instead of the nanoparticles used herein. It would also be of great interest to investigate if it is possible to use a current to induce the structure helical structure in the peptide.
This research will hopefully give us an insight into the interaction between surfaces and biomolecules, and how this might affect the everyday research in systems related to these interactions. It might also help us understand the forces driving protein folding.
The ability to make nanostructures on surfaces is extremely important in the ever-growing field of nanoscience, and with a system like this it could in the future be possible to use it as an organic memory, were, for example, the random coil represents a 0 and the helix makes a 1.
