WO2005005464A2

WO2005005464A2 - Peptide-based linker

Info

Publication number: WO2005005464A2
Application number: PCT/GB2004/002939
Authority: WO
Inventors: Derek Woolfson; Maxim Gennadievich Ryadnov; Bülent CEYHAN; Christof Niemeyer
Original assignee: University of Sussex
Current assignee: University of Sussex
Priority date: 2003-07-08
Filing date: 2004-07-08
Publication date: 2005-01-20
Anticipated expiration: 2006-01-08
Also published as: GB0315993D0; WO2005005464A3

Abstract

The invention relates to a self-assembling peptide linker comprising a full size peptide which templates the assembly of smaller peptides.

Description

Peptide-Based Linker

The present invention relates to peptide linkers capable of linking components together. The present invention also relates to methods of forming peptide linkers and linking components together.

Biomolecular recognition systems, such as nucleic-acid hybridization and protein-ligand interactions, are being utilized in biosensor (1) and microarray technology (27). There is also interest in using biomolecules as components in supramolecular self-assembly to generate novel components for the development of advanced functional materials and molecular electronics (8, 9). A key driver in these fields is the demand for smaller, cheaper and more-complex components. It is widely accepted that such improvements will not be fully realized using conventional "top-down" methods (e.g. photolithography), but will require fundamentally new approaches for the fabrication of electronic and other components (8). Consequently, complementary "bottom-up" technologies are being explored as potential ways for fabricating nanometer-sized devices. These are based on the self-assembly of small molecular building blocks to form larger functional elements. Biomolecules have been established as effective tools for mediating supramolecular assembly of molecular and collodial components. In particular, due to the specificity of Watson-Crick base-pairing, DNA oligonucleotides having unique and predictable self-recognition capabilities that make them promising starting materials for constructing well-defined nanostructures and assemblies of metal and semiconductor nanoparticles (8, 13, 14).

One concept for using biomolecules to assemble nanoparticles is illustrated in Fig 1A, which is an illustration of the assembly of nanoparticles using biomolecular linkers. Nanoparticles are functionalized with individual recognition groups that are complementary for a separate molecular linker. Addition of the linker drives assembly of the particles to form extended networks, which, in some cases, grow to macroscopic materials. Mirkin's group has pioneered this approach using oligonucleotide-based recognition motifs and has successfully applied it to solution-phase assembly in nanoparticles, (15-21) and the directed immobilization of nanoparticles on solid substrates, which allows for the fast and sensitive detection of nucleic acid analytes (22-25).

As the relationships between the primary and three-dimensional structures of peptides and proteins are less clear than those for DNA, it is not straightforward to envisage how polypeptides might be employed in the above scheme. Nonetheless, some similar strategies have been explored using protein-based recognition systems, in particular, antibody-antigen (26) and streptavidin-biotin (27-29) interactions. Directed molecular evolution strategies have also been applied to develop peptides that recognize the inorganic surfaces of semiconductors and metal nanoparticles, thus allowing the mediated growth of hybrid supramolecular networks. However, these approaches have drawbacks, and none provides a general means to achieve efficient biomimetic assembly of molecular and colloidal components. Exploring novel linker systems would enrich the kit of methodologies leading to determining easily reproducible principles for biomolecular-directed assembly. In addition, peptide-based linking systems introduce the possibility of recombinant protein production and, with it, the possibility of producing functional protein fusions as building blocks for new bioinspired materials.

The present invention provides a self-assembling peptide linker comprising a peptide belt and a plurality of peptide braces, the peptide belt comprising a plurality of peptide monomer units arranged as a strand, and each of the peptide braces comprising a plurality of peptide monomer units arranged as a strand, wherein the plurality of monomer units of the peptide braces are aligned with the plurality of monomer units in the peptide belt to form a substantially blunt ended coiled coil structure.

Such a linker allows easy assembly of molecular or colloidal components for use in various aspects of nanotechnology. In particular, the linker allows components to be linked together. The peptide linker is self-assembling, meaning that its elements will associate together to form a stable structure. The linker comprises a peptide belt, which acts as a template for the assembly of the linker, bringing together the plurality of peptide braces. When mixed under appropriate conditions the peptide belt and peptide braces will automatically assemble to form the linker. Appropriate conditions for allowing the self-assembly of such peptides are well known in the art.

The peptide belt and peptide braces are peptides that together form a coiled coil. Coiled coils are α-helical bundles with strands arranged in parallel, antiparallel or mixed topologies. Preferably the linker forms a leucine zipper. A leucine zipper is a type of coiled coil with two parallel strands. Coiled coil structures, and especially leucine zippers, are well known to those skilled in the art.

The peptide linker provided by the present invention introduces a new concept over what has gone before, namely that one full size peptide (belt) can template the assembly of smaller peptides (braces). In the absence of the belt, the braces do not associate. This is different from standard coiled-coil assembly, and has several advantages, including the fact that assembly of peptide can be initiated simply by adding the belts to the braces.

The term "peptide" as used herein refers to a polymer of amino acids and does not refer to a specific length of the product; thus, peptides, oligopeptides and proteins are included within the term "peptide". Equally the terms "protein" and "oligopeptide" are deemed to have the same meaning. The terms also do not exclude post-expression modifications of the peptide, for example, glycosylations, acetylations and phosphorylations. Included in the definition are peptides containing one or more analogs of an amino acid (including for example, unnatural amino acids), peptides with substituted linkages, as well as other modifications known in the art, both naturally occurring and synthesised.

The peptide belt and peptide braces are made up of peptide monomer units, which comprise a series of amino acids selected to bring about the self assembly of the peptide belt and peptide braces into the linker. Preferably the monomer units of the peptide belt and peptide braces comprise a heptad repeat motif (abcdefg), which is a 7 - residue repeat of hydrophobic (H) and polar (P) residues, arranged in the pattern HPPHPPP. The residues in the repeat motif in the peptide belt are chosen so as to attract residues in the repeat motif in the peptide braces. Alternatively the monomer units may comprise a hendecad repeat motif (abcdefghijk). Preferably the monomer units consist of the heptad or hendecad repeat motif. Such heptad and hendecad repeat motifs are well known to those skilled in the art and are described in WOO 1/21646.

The monomer units of the peptide braces are aligned with those in the peptide belt. This means that the first monomer unit in the peptide belt is aligned with the first monomer unit in one of the peptide braces, and the last monomer unit in the peptide belt is aligned with the last monomer unit in another peptide brace. Also, the first amino acid in each monomer unit in the peptide belt is aligned with the first amino acid in each of the corresponding monomer units in the peptide braces. Where heptad or hendecad repeat motifs are used, an amino acid in one position of the heptad (or hendecad) in the peptide belt is in line with an amino acid in the corresponding position in the heptad (or hendecad) in the peptide braces, a with a, b with b, c with c etc.

The linker is substantially blunt ended, which means that the ends of the peptide belt and peptide braces at each end of the linker are substantially flush with each other. That is to say, the peptide belt and peptide braces do not substantially over hang one another. If there is an overhang it is so small that the free end is not capable of interacting with the free end of another linker, thereby preventing the linkers of the present invention joining together to form an extended structure. Preferably there is less than 4 amino acids overhanging, more preferably less than 3 amino acids. Most preferably there is no overhang. This prevents the linkers from associating together.

The sum of the length of the plurality of braces is preferably substantially equal to the length of the belt. Preferably the belt and the sum of the braces differ in length by no more than 4 amino acids, more preferably no more than 2 amino acids. When a heptad repeat motif is used, the residues at positions a, d, e and g preferably constitute the interface between the peptide belt and the peptide braces.

When either a heptad repeat motif, or a hendecad repeat motif is used, the amino acids at the a positions are preferably either isoleucine or asparagine. More preferably all but one of the residues at the a position in the peptide belt are isoleucine, and the remainder is asparagine. Where there is an asparagine in an a position in the peptide belt, there is preferably also an asparagine at the "corresponding position" in at least one of the peptide braces. "Corresponding position" means at the position in the peptide brace aligned with the asparagine in the peptide belt when the linker is formed. The remainder of a positions are preferably isoleucine. The inclusion of a pair of asparagine residues at corresponding a positions allows the interaction of the peptide braces with the peptide belt to be distinguished. It also aids in positioning the peptide braces, relative to the peptide belt, because the peptide belt and peptide braces will associate so that the asparagine residues are aligned.

When either a heptad repeat motif, or a hendecad repeat motif is used the e and g positions in the peptide belt are preferably negatively charged residues, more preferably glutamic acids. The e and g positions in the peptide braces are preferably basic residues, more preferably lysine. This helps to prevent autonomous folding of the peptide belt and peptide braces and encourages tertiary assembly.

When either a heptad repeat motif, or a hendecad repeat motif is used the amino acids in the b and c positions are preferably alanine, glutamic acid, glutamine or lysine. For example both b and c may be alanine, or b may be alanine, and c glutamic acid. The amino acids in the /positions are also preferably alanine, glutamic acid, glutamine or lysine, but with one position in each strand being a modified amino acid, such as a tyrosine chromophore.

It is preferable that the peptide belt is composed of from 4 to 10 units, preferably 6 to 8, most preferably 6 units. The peptide braces may or may not be equal in length. There are preferably 2 to 5 peptide braces, most preferably 2 peptide braces. In the case of 2 peptide braces, the peptide braces are preferably half the length of the peptide belt, and are equal in length. The peptide braces are preferably composed of between 2 and 5 units, more preferably 3 and 4, most preferably 3 units. The sum of the number of units in the peptide braces is preferably equal to the number of units in the peptide belt.

At a neutral pH the belt is preferably negatively charged, and the braces are preferably positively charged, or vice versa.

In a particularly preferred embodiment the linker of the present invention comprises a peptide belt and a first and a second peptide brace. The N terminus of the first peptide brace is preferably aligned with the N terminus of the peptide belt. A CysGlyGly tag is preferably added to the N terminus of the first peptide brace. The C terminus of the second peptide brace is preferably aligned with the C terminus of the peptide belt. A GlyGlyCys tag is preferably added to the C terminus of the second peptide brace. This allows derivatisation. Except for the C terminus of the first peptide brace, and the N terminus of the second peptide brace, the termini are preferably capped.

The peptide belt preferably has the sequence: IAALEKEIAALEQEIAALEKEIAALEYENAALEKEIAALEQE

The first peptide brace preferably has the sequence: IAALKQKIAALKQKIAALKYK

The second peptide brace preferably has the sequence: IAALKQKN-AALKQKIAALKYK

The linker allows components to be linked together, each component being attached to a separate peptide brace of the linker, or to the peptide belt, and the peptide belt and peptide braces together forming a reversible, stable structure. Preferably when linking components, the components are attached to separate peptide braces. The components may be, for example, molecules, such as molecules in solution, peptides or proteins and peptide bound cargo, such as pharmaceuticals or analytical reagents. The term "components" also encompasses surfaces, for example gold surfaces.

Further provided is a method of producing peptide linkers according to the invention comprising providing a mixture of peptide belt and peptide braces which associate to form a self-assembling peptide linker comprising a peptide belt and a plurality of peptide braces, the peptide belt comprising a plurality of peptide monomer units arranged as a strand, and each of the peptide braces comprising a plurality of peptide monomer units arranged as a strand, wherein the plurality of monomer units of the peptide braces are aligned with the plurality of monomer units in the peptide belt to form a substantially blunt ended coiled coil structure.

The peptide belt and peptide braces may be made by any standard procedure known to those skilled in the art, for example synthetically or recombinantly.

The invention also provides a kit for making a peptide linker according to the present invention comprising a peptide belt and peptide braces which associate to form the peptide linker according to the invention.

Also provided is a method of linking components comprising using the peptide linker according to the invention. Each of the components to be linked are separately coupled to, or include, the different elements of the linker (i.e. the peptide belt and the plurality of peptide braces). The peptide belt and peptide braces are then allowed to associate and form the linker, thereby linking the components together. Preferably each component is linked to a separate peptide brace. Preferably the components and linker are formed separately, and the method includes the step of coupling the components to the peptide belt or peptide braces, prior to formation of the linker. Alternatively the peptide belt or peptide braces may be made as part of the components. This is particularly true of protein components. The invention also provides a method of linking a protein to another component, which may or may not be another protein, wherein the protein is recombinantly formed and includes the peptide belts or the braces of a peptide linker of the invention. The method of linking the components is preferably performed in solution.

In particular, the invention provides a method of linking a component to a surface (i.e. a solid support) using the peptide linker according to the invention. Preferably the peptide belt is coupled to the surface, and the peptide braces are coupled to other components. On allowing the peptide belt and the peptide braces to associate, the components will be linked to the surface. The method can particularly be used for coupling components to gold surfaces, for use, for example, in biosensors.

The invention further provides a method of spacing components comprising coupling components to a peptide linker according to the invention. The size of the linker is accurately known, hence by attaching the components to the linker their position in relation to each other can be accurately predicted. The peptide linker according to the invention may also be used as a ruler or spacer, to measure distances in nanotechnology. One linker can be used to space two or three components, or a plurality of linkers can be used to space more components, for example linearly, or forming a grid. This can be achieved by coupling each component to more than one belt or brace, so that each component will be coupled to more than one linker. When the linkers assemble they form a network.

Further provided is a scaffold comprising a peptide linker according the invention. The peptide linker can be used as a bridge between components in order to form a molecular scaffold.

The peptide linker can also be used to form hydrogels.

The invention will now be described in detail, by way of example, with reference to the further drawings Fig. IB to 8 in which:

Figure 1 shows linker-mediated assembly of nanoparticles. Figure IB illustrates peptide linkers in accordance with the invention, showing preferred linear sequences of the belt (A) and braces (B_N and Be). Designed inter-peptide electrostatic interactions are depicted by double-headed arrows. The break between the BN and B_c peptides is indicated by scissors. In (C) the sequences are drawn out on a 3.5-residue-per-turn helical wheel to show how the interfacial a, d, e and g positions of the heptad repeats come together. The break between the B_N and Be peptides occurs between K ₂ and 1₂5, which are highlighted bold.

Figure 2 is an illustration of a peptide assembly including peptide linkers in accordance with the invention probed by CD spectroscopy. (A) Spectra for the individual first strand (broken line), a 1:1 mixture of the second and third strands (dotted line) and the 1:1:1 mixture of all three strands (solid discs). The open discs show the spectrum of the latter after thermal unfolding and cooling. (B) Spectra for one of the two-component mixtures, namely BN plus the first strand in 1 :1 (dotted line) and 2:1 (dashed line) ratios, solid discs are as in (A). Samples were 100 μM in each peptide, 10 mM MOPS, 1 mM DTT, pH7 and at 5°C.

Figure 3 is a thermal unfolding image of a 20μM peptide linker in accordance with the invention. Open discs show the raw CD data and the broken line gives the first derivative.

Figure 4 is an illustration of pH stability of a 20μM peptide linker in accordance with the invention. (A) CD spectra for 1:1:1 mixtures of the peptides recorded at pH 5.4 (solid line), 7 (crosses) and 8.6 (dashed line). (B) [θ]₂₂₂ versus pH.

Figure 5 is an illustration of on-surface assembly of peptide linkers in accordance with the invention followed by SPR (A) shows step-wise assembly of a peptide linker on a sensor chip starting with the coupling of Be followed by the addition of A and then B_N*. (B) shows coupling of BN followed by washing and addition of a 1:1 mix of A + Be*. (C) shows peptide-concentration dependencies of the coupling of the cysteine-containing braces (open discs) and subsequent complex formation using the sequential (solid triangles for A, and open triangles for BN) and the alternative methods (solid discs). In the schematic representations of the coiled-coil assemblies N represents asparagines. Conditions: except for the experiments shown in panel D where peptide concentrations were varied, all peptide injections were 50 μL of 20 μM of each peptide in HEPES buffered saline at pH 7.4 and room temperature.

Figure 6 shows a UV-visible spectra for the 1 :1 BN^AU:B_C ^AU mixture before (solid line) and after (open circles) adding a stoichiometric amount of peptide A, the first strand.

Figure 7 shows direct observation of peptide-mediated nanoparticle assembly. TEM images of ordered three-dimensional (A, B) and two-dimensional (C) gold-nanoparticles networks formed by mixing BN^AU, B_C ^AU and peptide A in a 1 :1 :1 ratio. D, a typical colloidal material from a mixture of the linker with one brace (BN^AU or Bc ^Au) in 1:1 or 1:2 ratios.

Figure 8 shows folding of the nanoparticle-bound peptides. CD spectra for the following equimolar mixtures: B_N ^Au + B c^Au (dotted line); B _N ^{A U} + A (dashed line); B_N ^Au + Be ^Au + A (solid line); B_N + B_c ^Au + A (crosses).

Examples

Here we report the rational design of a peptide recognition system based on a three-component coiled-coil assembly, which we refer to as "belt and braces", (Fig. IB). In terms of peptide and protein structure, coiled-coil assemblies are relatively simple and sequence-to-structure relationships are available that permit their prediction and design. Briefly, coiled coils are bundles of between two and five amphiphatic helices, which assemble through their hydrophobic faces, Fig IC. (33, 34) The nature of the residues at the interface determines coiled-coil oligomer state, helix orientation and partner selection. (33 -40) The simplest and best understood coiled coils are the two-stranded, parallel leucine zippers. Based on our understanding of leucine zippers, we designed a belt peptide to template bringing together two brace peptides. We demonstrate specific assembly of the ternary complex as well as its utility in peptide-directed immobilization at surfaces and the assembly of nanoparticles. Example 1 Peptide Design

Our aim was to engineer a peptide-based linker system of three components: a belt (A, Figure IB), which would act as the template for the assembly of two braces (BN and Be). Others have introduced a similar approach to make self-replicating peptides. Coiled coils, specifically leucine-zipper motifs, are ideal candidates for such designs because principles are available that link coiled-coil sequence and structure (33-40). We built on our own experience in leucme-zipper design (42-45) to engineer the "belt-and-braces" system. This process involved both engineering towards the desired structure (positive design) and away from alternate structures (negative design) (42-46); for instance, the possibilities for autonomous folding of the individual peptide components, or non-productive co-assembly of pairs of components had to be minimized. The resulting design had the following features, Figures 1, B & C: first, the sequences contained heptad repeats, abcdefg, typical of canonical coiled coils. In this notation, residues at a and d and at e and g contribute to the helix-helix interface, as shown in Figure IC, and are usually hydrophobic and charged, respectively. Four or more contiguous repeats usually produce stable assemblies. We settled on a six-heptad design for the belt and braces: the belt comprised six heptads. whilst the braces had^* three each. Second, to direct a parallel, two-helix (leucine-zipper) structure all but one of the a positions was made isoleucine and all of the d were made leucine.(36-37) Third, to prevent folding and to encourage ternary assembly, the e and g positions of the belt were all made charged (glutamic acid), whilst the corresponding positions of the braces were all made basic (lysine). Fourth, to distinguish the interactions of the two braces and the belt, the central a position of the C-terminal brace (Be) was made asparagine, as was its complement, the fifth a position in the belt; though destabilizing, the resulting asparagine-asparagine interaction is highly specific (40, 43, 48, 49). Fifth, the remaining b, c and /positions were made combinations of alanine, glutamic acid, and lysine; one /position in each peptide was reserved for a tyrosine chromophore. Sixth, to allow derivatisation, CysGlyGly and GlyGlyCys tags were added to the N and C-termini of B_N and Be, respectively. Finally, except for the C and N-termini of BN and Be, respectively, the termini of all the peptides were capped.

Experimental Section

Material and Methods.

In the following experiments the following materials and methods were used. A Pioneer Peptide Synthesis System (PE Applied Biosystems, CA, USA) was used for peptide assembly. All reagents and resins were purchased from Applied Biosystems (Warrington, UK) or CN Biosciences (Nottingham, UK). Colloidal gold nanoparticles (15 nm) were from ICN Biomedicals Gmbh (Eschwege, Germany). Analytical and semi-preparative gradient RP-HPLC was performed on a JASCO HPLC system (Model PU-980- Tokyo, Japan) using Vydac Cis analytical (5 pm, 4.6 mm i.d. x 250 mm) and semi-preparative (5 μm, 10 mm i.d. x 250 mm) columns. Both analytical and semi-preparative runs used a 20-40% B gradient over 45 min at 4.7 mL/min (semi-preparative) and 1 mL/min (analytical) where buffer A was 5% aqueous CH₃CN, 0.1% TFA, and buffer B was 95% aqueous CH₃CN, 0.085% TFA. Mass spectra were recorded on a TofSpec E Matrix Assisted Laser Desorption Ionization (MALDI) spectrometer (Micromass Ltd, Manchester, UK). Analytical ultracentrifugation was performed using a Beckman Optima XL-I analytical ultracentrifuge fitted with an An-60 Ti rotor. CD experiments were conducted on a JASCO J-715 spectropolarimeter fitted with a Peltier temperature controller (Tokyo, Japan). SPR experiments were carried out on BIACORE 2000 instrument (Biacore AB, Uppsala, Sweden). TEM data were acquired using Hitachi 7 100 transmission electron microscope (Tokyo, Japan), fitted with a charge-coupled device camera from Digital Pixel Co. Ltd. (Brighton, UK) and software from Kinetic Imaging Ltd. (Liverpool, UK).

Peptide synthesis

Peptide synthesis was carried out by the combination of standard Fmoc/tBu solid phase protocols with TBTU/DIPEA as coupling reagents on a PEG-PS-resin for carboxyl-free peptides and using a PAL linker for peptide amides. A 95% TFA mixture (95:2.5:2.5 TFA/water/TIS) was used as the post-synthesis cleavage cocktail for non-cysteine peptides. For the deprotection of the cysteine-containing peptides another 93.5% TFA mixture (93.5:2.5:2.5:1.5 TFA water/EDT/TIS) was used. The purification of all peptides used semi-preparative RP-HPLC, and the purities were confirmed by analytical RP-HPLC.

Mass spectrometry.

Peptide identities were verified by MALDI-Tof mass spectrometry with a-cyano-4-hydroxycinnamic acid as the matrix: peptide A [M+H]+ m/z 4624.2 (calc), 4625.4 (found), [M+2H]z+ m z 2313 (found); peptide B_N [M+HJ+ m/z 2570 (calc),

2571.2 (found), [M+K]+ m/z 2609.7 (found); peptide B_N* [M+H]+ m/z 2354 (calc), 2354.8 (found); peptide Be [M+H]+ m/z 2528.1 (calc), 2529.3 (found), [M+Na]+ m/z 2551.1 (found), [M+KJ+ m/z 2567.2 (found); peptide B_c* [M+H]+ m/z 2311.9 (calc),

2312.3 (found), [M+Na]+ m/z 2334.8 (found); peptide C [M+H]+ m/z 2826.3 (calc), 2828.3 (found), [M+2H]z+ m z 1413.2 (found), [M+K]+ m/z 2865.4 (found).

Sedimentation equilibrium experiments.

Sedimentation equilibrium experiments were conducted at 5°C. 100 μL samples of 100 μM peptide A or mixtures of the braces, and the belt plus both braces, with initial A₂₈o values being of 0.15, 0.147 and 0.144, respectively, in the 1.2 cni path length cells used, were buffered to pH 7 with 10 mM MOPS containing 1 mM DTT and 100 m.M sodium chloride. Samples were equilibrated for 48 h at 30000, 37500 and 55000 rpm. Sedimentation equilibrium curves were measured by the absorbance at 28C resulting data were fitted simultaneously using routines in the Beckman Optima )M-A/XL-I da-a software (version 4.0). The density of the buffer at 5°C was taken as 1.005 mg/mL. Based on the amino acid composition the averaged partial specific volume for the peptides was calculated to be 0.75 mL/mg for peptide A, 0.774 mL/mg for B_N/B_C, and 0.77 mL/mg for the equimolar mixture of A, BN and Be.

Circular dichroism spectroscopy. All data for peptide samples prepared in 10 mM MOPS (pH 7) were collected in 1-mm quartz cuvettes. Data points for CD spectra were recorded at 1-nm intervals using a 1-nm bandwidth and 4-16-s response times. After baseline correction, ellipticities in mdeg were converted to molar ellipticities (deg cm² dmol res^"1) by normalizing for the concentration of peptide bonds. Data points for the thermal unfolding curves were recorded through l°C/min ramps using a 2-nm bandwidth, averaging the signal for 8s at 1 °C intervals. Data for pH experiments were obtained for peptide samples in 10 mM EPPS (basic pH) and MES (acidic pH) buffers.

Surface Plasmon Resonance (SPR).

SPR sensograms were recorded at 20°C on a BIACORE 2000 instrument using standard procedures and buffers (HBS running buffer (10 mM HEPES, 0.15 M NaCl, 0.005% Surfactant P20, 50 μM EDTA (pH 7.4))). Peptide solutions were 20 μM unless otherwise stated. Bare gold sensor chips were used throughout the studies. These allowed a simple and reversible strategy for assembling thiol-containing braces followed by belt-plus-brace mixture on solid surfaces.

Preparation of peptide-nanoparticle conjugates.

Gold nanoparticle conjugates were synthesized by treating 15 nm gold nanoparticles (1 ml of a 1 nM aqueous solution) separately with peptides Be and BN (25 ycl of 3 mM solutions, to give final peptide concentrations of 73 μM). After standing overnight the solution was centrifuged at 14000 rpm. The supernatant was removed and the red sediment washed twice and resuspended in water. «65 - 80% of the original Au nanoparticle concentration was recovered.

UV-visible spectroscopy.

UV/Nis spectra were recorded on a Hitachi U3000 spectrometer. Quantification of the Au-peptide conjugates and spectral measurements of the aggregation experiments were carried out at room temperature in water.

Transmission electron microscopy. Unless stated otherwise, samples of Au-peptide conjugates were mixed 1:1 with peptide A (20-100 μM) in filtered MOPS buffer (10 mM, pH 7) and incubated at 20 °C for 30-60 min. Aliquots of the mixtures were applied to carbon-coated copper specimen grids (Agar Scientific Ltd, Stansted, UK), and dried with filter paper at room temperature. Grids were examined by a TEM at 100 kN and digital images (800 X 1200 pixel) recorded for analysis. Separations between neighbouring particles were measured and averaged over 300 pairs, the standard deviation of these measurements was 0.34 nm). The sizes of nanoparticles (15 nm) was confirmed with a standard deviation of lnm.

Solution-phase assembly of belt and braces.

Circular dichroism (CD) spectroscopy provides a convenient probe of α-helical structure in leucine-zipper systems (50). Consistent with the design concept, none the individual belt-and-braces peptides nor the combination of the two braces showed appreciable α-helix in solution by CD spectroscopy, (Figure 2A). However, equimolar mixtures of all three peptides gave CD spectra indicative of considerable α-helix formation, Figure 2A. Furthermore, as expected for an oligomerizing system, the amount of helix was concentration dependent: based on the CD signal at 222 nm ([θ]222), the percent helix for samples of 20 μM, 100 μM and 200 μM of each peptide were «70%, ∞80% and «85%, respectively; these values were consistent with TFE-induced helix formation, which gave a benchmark for the upper limit of the [θ]₂₂ in our system of «-32,000 deg cm² dmol^"1 (45, 52).

Thermal denaturation of leucine-zipper peptides as followed by the loss of [θ]₂₂₂ with increasing temperature provides a measure of stability and cooperatively of assembly. The thermal unfolding of 1 : 1 : 1 mixtures of the belt-and-braces peptides gave sigmoidal unfolding curves as expected for a unique, cooperatively folded structure, as seen in Figure 3. Importantly, low-temperature CD spectra recorded before and after thermal unfolding were very similar, as seen in Figure 2A, indicating almost complete reversibility of unfolding. The first derivatives of the unfolding curves were dominated by a single peak revealing transition midpoints (T ) of 60 ± 2°C, (Figure 3). Unexpectedly, however, the T_M was similar at 20 and 100 μM peptide, and the peaks had some structure possibly suggestive of partial unfolding or fraying of the helical assemblies. However, this behaviour was much more marked for the binary mixtures - for example, A + B_N (Figure IS, supporting information). These gave approximately linear thermal denaturation curves, which probably comprised several overlapping transitions, and were consistent with the formation of two or more non-specific complexes.

The oligomerisation states of the belt-and-braces peptides were probed by sedimentation equilibrium analysis in an analytical ultracentrifuge. In these experiments all samples were 100 μM in each peptide, 10 mM MOPS, 1 mM DTT, pH 7 and at 5°C sedimentation equilibrium data were fitted assuming a single ideal species. The returned molecular weights for the belt alone and for the 1 :1 mixture of the braces were consistent with the design and the foregoing CD data; that is, all peptides behaved as monomers and fitted the models well; the experimental molecular weights were 4875 Da (95% confidence limits 4592 and 5157) for the belt, and 2690 Da (2518 and 2857) for the braces, respectively, compared with the calculated relative molecular masses of 4624 for the belt, and 2570 and 2528 for the braces. The molecular weight returned for the 1 :1 :1 mixture, 9204 Da (9020 and 9387), was close to that expected for the 1:1 :1 ternary complex (9722 Da). However, the residual signal (experimental data minus the fitted curve) showed a systematic error typical of an associating system, though the application of more-complex association models was not deemed appropriate. Thus, using analytical ultracentrifugation the component peptides alone behaved as monomers; and, although the experimentally calculated mass for the 1 :1 :1 mixture was lower than expected for a fully folded complex, it was consistent with complete folding observed by CD.

As the belt-and-braces system comprises oppositely charged peptide strands - at neutral pH, the belt is negatively charged and the complementary braces are positive - we probed the α-helical content of the 20 μM 1:1 :1 complex in the pH range 5.4 and 8.6, Figure 4. CD spectra in this range all showed a α-helix, Figure 4 A, whereas spectra of the individual peptides as well as the 1 :1 mixture of indicated random coil. However, the helical content of the 1 :1:1 mixtures fell off rapidly outside the pH range 6.8 - 7.4,

Figure 4B.

Unlike in DNA assembly, where base pairing specifies partner selection precisely, principles for heterodimeric leucine-zipper association are less clear-cut. As a result, unintended and promiscuous peptide-peptide interactions may occur (42). To address this issue in the belt-and-braces system, we measured the α-helical content of the two belt-plus-brace binary combinations (A + BN and A + Be). Both gave helical CD spectra, but the α-helix content was < 60% at 100 μM. Although slightly more helix could be induced with 2:1 brace:belt ratios, the helical content never matched that observed for the 1:1:1 mixtures of the belt-and-braces peptides at similar concentrations, Figure 2B. Furthermore, unlike the sigmoidal thermal unfolding curve observed for the complete ternary mixture, Figure 3, neither of the binary mixtures showed any sign of cooperative unfolding. Interestingly, the A + B_N combinations was more helical than the A + Be mixtures (data not shown). This fits with the design principles as the A: Be interaction was engineered to be less stable, but more specific, than A:BN through the inclusion of the asparagine residues at one of the a layers of the A:Bc interface. Thus, we suspect that the A:BN interaction is both more stable and more promiscuous than the A:Bc association; Be can only pair with A at one position (covering heptads 4, 5 and 6), whereas BN can potentially pair at two positions (covering heptads 1, 2 and 3 and 2, 3 and 4). As mentioned above, the first derivative of thermal unfolding curve for A + B_N mixture provides support for this contention.

Example 2 Assembly of belt and braces on surfaces.

The reversible assembly of the belt-and-braces peptides into a complex offers a novel route to decorating solid surfaces with molecular and colloidal components (11, 22-25,

53, 54). Gold surfaces provide a convenient and potentially useful substrate to test this as the braces each had terminal thiol groups, which are readily chemisorbed onto gold from aqueous solutions. Furthermore, employing gold chips in Biacore allows binding and complex formation to be monitored in real time through surface plasmon resonance (SPR) measurements (55). For these reasons, we derivatized bare gold Biacore chips with the thiol-containing brace peptides, and to use the resulting functionalized surfaces to pull down the belt peptide in combination with non-thiol containing variants of the second brace peptide. To this end, we remade the brace peptides without the terminal cysteine residues, which we refer to as B_N* and Be*.

Figure 5 shows typical SPR experiments in which the individual braces, and their binary combinations with the belt were passed over a Biacore sensor chip and the binding monitored. These were continuous-flow experiments, in which the instrument was first equilibrated with standard running buffer. Peptide-containing solutions were then injected, and after some defined time flow of the running buffer was resumed to wash away any unbound material. As changes in refractive index of the different buffers and solutions affect the SPR signal, only equilibrium signals - for example, the starting (t_s) and final (t_F) values - can be compared reliably. These signals, quoted in resonance units (RU), were used to gauge the mass of material bound to the surfaces and the percentage of complex formation on the chips.

In general, two types of SPR experiment can be envisaged. Either, the complex is assembled on a surface by sequential addition of the component peptides; that is, brace, followed by belt, followed by second brace. Or, one brace is laid down on the surface prior to the addition of a binary mixture of the other brace and the belt. We compared both methods, Figures 5 A&B.

For the sequential experiment, a gold chip was treated with Be, followed by A and then BN*, Figure 5A. The surface was washed with running buffer between each peptide addition. The initial binding of Be was good, «140 RU. Although it has been reported that conformational changes can affect SPR signals (53), it is generally accepted that the change in RU reflects the mass of material on the surface. As a rule of thumb, 1000 RU corresponds to approximately 1 ng of material (56). Thus, the 1.40 RU corresponded to «50 fmoles of bound brace. However, the subsequent binding of A (an additional =20 RU) and then B_N* (a further «10 RU) were poor, Figure 5 A: based on the 140 RU of binding for B_c, the maximum theoretical signal for complete B_C:A:B_N* complex formation is = 4*140 = 560 RU, and we estimated that only « 7% [((170-140)/560-140)) * 100] of B_c was complexed. Presumably, the weak B_C:A interaction did not sequester enough of A at the surface (and/or did not survive the washing step) to allow recruitment of B_N to complete complex formation. However, the reverse experiment - that is, with B_N laid down first followed by addition of A and then Be* - gave similar results.

In the second approach, Figure 5B, the surface was initially modified with BN, washed and then treated with a pre-mixed 1:1 solution of A and Be*. After washing, a steady baseline of =230 RU was achieved. This represented an additional 90 RU of binding, which corresponded to =21 % complexation of the sequestered brace. Thus, addition of a binary mix of belt and brace to a surface-bound complementary brace improved complex formation over the sequential assembly method. A similar final intensity 0240 RU) was achieved in the reversed experiment in which the chip was treated with the A + Be combination followed by BN*.

These experiments were done using solutions 20 μM in each peptide. At higher concentrations the binding capacity of the cysteine-containing braces was considerably improved, Figure 5C; for example, >900 RU at 1 mM peptide (i.e. 320 fmoles of brace). Sequential assembly did not improve with increased peptide concentrations. Although the addition of 1 : 1 beltbrace* mixtures did show increased absolute amounts of complex formation these appeared to saturate and fall short of 100% complexation of the sequestered brace.

To assess non-specific peptide binding due to protein adhesion and electrostatic interactions between oppositely charged peptides two control experiments were carried out: (1) treatment of a Bc-covered chip by similarly charged BN* showed no binding; (2) a positively charged control coiled-coil peptide -C, Ac- CGGYGAQIAALKQQNAALKQQIAALKQ-NH, - was engineered as a "false brace" unable to co-assemble with the belt and attached to the surface. As expected., C failed to recruit A + Be* mixtures.

The system showed good reversibility: once assembled on the gold surface the BN:BC:A complex was readily disassembled and removed by treating with 1 M guanidine hydrochloride. The residual signals were as expected for the binding of the brace peptide(s) alone. These surfaces could either be reused for another round of assembly, or desorbed completely from the surface using mercaptoethanol(57).

Example 3

Peptide-directed nanoparticle assembly.

If fully folded, the belt-and-braces coiled coil would span 6 to 7 nm. Thus, the system has potential as a self-assembling nanoscale linker of defined length. We tested this through the assembly of gold nanoparticles to extend the concept of DNA-based nanoparticles assembly, Figure 1A.

The brace peptides, BN and Bc, were separately coupled to 15-nm colloidal gold particles to give BN^AU and Bc^Au. After removal of excess, unbound peptide, the derivatized particles were combined. The characteristic red color of the gold suspensions did not change either after peptide coupling, or mixing. By UV-Vis spectroscopy, the conjugates showed no significant changes in gold-absorbance spectra compared with unmodified gold particles (data not shown). This is consistent with no assembly taking place and the designs of the belt-and-braces system. In addition, flocculation assays revealed that both of the Au-peptide conjugates were more stable to increased salt (up to 1 M NaCl) than the bare An colloids. This behavior is possibly explained in that the brace peptides are positively charged and, so, inhibit particle aggregation and growth. By contrast, addition of the belt peptide to the mixture of brace-peptide-derivatized gold nanoparticles caused a color change and a precipitation of aggregated particles, which settled in the reaction vessel. Moreover, the surface plasmon resonance absorption band associated with the gold nanoparticles for this sample was broadened and red shifted (from 529 to 545 nm) compared with mixtures lacking the belt peptide, Figure 6. UN-Nis spectra for just one type of derivatized gold particle, either B_Ν ^Au or B ^Au combined with the belt did show a shift, but this was smaller (to 536 nm) and five-times less intense than with the complete belt-and-braces system. In summary, these results concur with the CD and SPR data and indicate that though there were some promiscuous belt-brace interactions the complete A + BN + Be system produced the most competent and stable assemblies.

Nanoparticle assemblies were visualized directly by transmission electron microscopy (TEM), Figure 7. As seen by others working with colloidal precipitates, we observed both 3-D and 2-D networks, Figures 7 A, and B & C, respectively. The former could not be analyzed with any certainty, but the 2-D networks showed uniformly separated particles at distances of 7.22 ± 0.34 nm. These measurements are consistent with a folded six-heptad coiled-coil linker. The networks were found exclusively in preparations containing all three belt-and-braces components, B_N ^AU, B_c ^Au and peptide A: under similar conditions, bare gold colloids displayed only closely aggregated particles without distinct spacings indicative of non-specific particle-growth processes. Similarly, in the mixtures of peptide A with BN^AU or Bc^Au single particles and non-uniform aggreagtes dominated, Figure 7D.

Further CD experiments tested if the belt-and-braces peptides were fully folded in the presence of the gold nanoparticles, Figure 8: in the absence of the belt peptide, a mixture of B_N ^AU and Bc^Au gave a random coil signal as expected; a mixture of BN^AU + A was «50% helical; whereas mixtures with all three peptide components were » 70% helical or more. Interestingly, the BN^AU + Bc^Au + A mixture was «10% less helical than the B + B_c ^Au+ A sample. It is unlikely that this loss in helix is due to unwinding of the belt-and-braces coiled coil; more likely, the bulky nanoparticles reduce complex formation to a certain extent and/or this effect is due to light-scattering commonly observed for particulate systems. Summary

To summarize, we describe a novel, self-assembling, nanoscale, peptide-based linker. The system comprises three leucine-zipper peptides of de novo design: the longer peptide, A or "the belt", acts as a template for the co-assembly of two half-sized peptides, B_N and Be, also termed "the braces". The basic features of the design were confirmed in solution by CD spectroscopy and analytical ultracentrifugation, which indicated a predominantly a-helical, ternary assembly. These assemblies were thermally stable and unfolded reversibly. Although specific positive and negative features were included in the design, some promiscuous belt-brace interactions were observed. However, these were not cooperatively folded structures, and were thermally labile. Using SPR experiments in Biacore, we demonstrated assembly of the complex on gold surfaces. This was best effected by first coupling one of the brace peptides to the surface via a thiol group, followed by co-assembly with a binary mixture of the belt and the other brace; that is, rather than by sequentially adding the component peptides to the surface.

The utility of the system in bringing together peptide-bound cargo was demonstrated using colloidal gold nanoparticles. The braces were used to derivatize the particles, which were then brought togther to form nanoscale networks by adding the belt. TEM images of these networks revealed that the particles were regularly separated by «7 nm, fully consistent with the six-heptad design of the belt-and-braces coiled coil. In accord with this, CD experiments confirmed that the majority of the peptide linkers were folded when bound to gold nanoparticles.

To our knowledge, belt-and-braces is a novel concept in coiled-coil assembly, and the first example of employing rationally designed peptides to guide nanoparticle assembly, which, hitherto, has been achieved using DNA-based linkers of some description.(18, 12-25) Though the field of DNA-directed assembly is more mature at present, peptides potentially offer some advantages over DNA in such applications. Specifically, synthetic genes for peptide sequences can be synthesized and cloned for use in recombinant DNA technologies. In this case, for protein cargos the braces need not be added chemically, but brace-protein fusions could be engineered recombinantly. This possibility may unlock exciting new routes to enable the generation of novel and functional protein-based supramolecular and nanostructured assemblies and biosensors (58).

All documents cited are incorporated herein by reference.

References

1) Frontiers in Biosensors: Fundamental Aspects; Scheller, F. W., Schaubert, F., Fredrowitz, J. (Eds.) Birkhduser Nerlag, 1997.

(2) Holloway, A. J.; Nan Laar, R. K.; Tothill, R. W.; Bowtell, D. D. Nat Genet 2002, 32 Suppl 2,481-489.

(3) Pirrung, M. C. Angew. Chem. Int. Ed. 2002,41,1276-1289.

(4) Mirzabekov, A.; Kolchinsky, A. Curr Opin Chem Biol 2002,6,10-15. (5) Templin, M. F.; StoU, D.; Schrenk, M.; Traub, P. C; Nohringer, C. F.; Joos, T. O. Trends Biotechnol 2002, 20,160-166.

(6) Lam, K. S.; Renil, M. Curr Opin Chem Biol 2002, 6,353-358.

(7) Winssinger, Ν.; Ficarro, S.; Schultz, P. G.; Harris, J. L. Proc NatlAcadSci USA 2002, 99, 11139-11144 (8) Νiemeyer, C. M. Angew. Chem. Int. Ed. 2001,40,4128-4158.

(9) Νiemeyer, C. M. Science 2002,291,62-63.

(10) Moore, G. E. Electronics 1965, 38, 114-116.

(11) Whitesides, G. M.; Mathias, J. P.; Seto, C. T. Science 1991, 254, 1312-1319.

(12) Seeman, Ν. C. Nano Lett. 2001, 1, 22-26. (13) Mirkin, C. A. Inorg. Chem. 2000,39,2258-2272.

(14) Gerion, D.; Parak, W. J.; Williamson, S. C; Zanchet, D.; Micheel, C. M.; Alivisatos, A. P. J. Am. Chem. Soc. 2002,124,7070-7074.

(15) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C; Storhoff, J. J. Nature 1996, 382, 607-609. (16) Elghanian, R.; Storhoff, J. J.; Mucic, R. C; Letsinger, R. L.; Mirkin, C. A. Science 1997, 211, 1078-1081

(17) Mucic, R. C; Storhoff, J. J.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc 1998, 120, 12674-12675.

(18) Storhoff, J. J.; Elghanian, R.; Mucic, R. C; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc 1998, 120, 1959-1964.

(19) Mitchell, G. P.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc 1999, 121, 8122-8123. (20) Reynolds, R. A.; Mirkin, C. A.; Letsinger, R. L. J Am. Chem. Soc 2000,122, 3795-3796.

(21) Storhoff, J. J.; Lazarides, A. A.; Mucic, R. C; Mirkin, C. A.; Letsinger, R. L.; Schatz, G. C. J. Chem. Soc 2000, 122, 4640-4650. (22) Taton, T. A.; Mucic, R. C; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc 2000, 122, 6305-6306.

(23) Taton, T. A.; Mirkin, C. A.; Letsinger, R. L. Science 2000, 289,1757-1760.

(24) Taton, T. A.; Lu, G.; Mirkin, C. A. J. Am. Chem. Soc 2001, 123, 5164-5165.

(25) Park, S. J.; Taton, T. A.; Mirkin, C. A. Science 2002, 295,1503-1506. (26) Shenton, W.; Davies, S. A.; Mann, S. Adv. Mater. 1999,11,449-452.

(27) Niemeyer, C. M.; Burger, W.; Peplies, J. Angew. Chem. Int. Ed. Engl. 1998,37,2265-2268.

(28) Connolly, S.; Fitzmaurice, O. Adv. Mater. 1999, 11, 1202-1205.

(29) Niemeyer, C. M.; Ceyhan, B. Angew. Chem. Int. Ed. 2001,40,3685-3688. (30) Lee, S. W.; Mao, C; Flynn, C. E.; Belcher, A. M. Science 2002,296,892-895.

(31) Whaley, S. R.; English, D. S.; Hu, E. L.; Barbara, P. F.; Belcher, A. M. Nature 2000,405,665-666.

(32) Brown, S. Nano Lett. 2001, 1, 391-394.

(33) Lupas, A. TIES 1996, 21, 375-382. (34) Walshaw, L; Woolfson, D. N. Protein Sci 2001,10,668-673.

(35) O'Shea, E.K.; Lumb, K.J.; Kim, P.S. Current Biology 1993, 3, 658-667.

(36) Harbury, P. B.; Zhang, T.; Kim, P. S.; Alber, T. Science 1993,262,1401-1407.

(37) Woolfson, D. N.; Alber, T. Protein Science 1995, 4, 1596-1607.

(38) Kim, B.M.; Oakley, M.G. JAm Chem Soc 2002,124,8237-8244. (39) Monera, O.D.; Kay, CM.; Hodges, R.S. Biochemistry 1994,33,3862-3871.

(40) Gonzalez, L; Woolfson, D. N; Alber, T. Nat. Struct. Biol. 1996, 3, 1011-1018.

(41) Saghatelian, A.; Yokobayashi, Y.; Soltani, K.; Ghadiri, M. R. Nature 2001, 409,797-801.

(42) Nautiyal, S.; Woolfson, D.N.; King, D.S.; Alber, T. Biochemistry 1995, 34, 11645-11651.

(43) Pandya, M. J.; Spooner, G. M.; Sunde, M.; Thorpe, J. R.; Roger, A.; Woolfson, D. N. Biochemistry 2000, 39, 8728-8734. (44) Ciani, B.; Hutchinson. E.G.,; Sessions R.B.; Woolfson, D.N. JBiol Chem 2002, 277, 10150-10155

(45) Hicks, M.R.; Holberton, D.V.; Kowalczyk, C; Woolfson, D.N. FoldDes 1997,2,149-158. (46) Beasley, J.R.; Hecht, M.H. JBiol Chem 1991, 212, 2031 -2034.

(47) Litowski, J.R.; Hodges, R.S. JPeptRes 2001, 58,477-492.

(48) O'Shea, E. K.; Klemm, J. D.; Kim, P. S.; Alber, T. Science 1991,254,539-544

(49) Lumb, K. J.; Kim, P.S. Biochemistry 1995,34,8642-8648.

(50) O'Shea, E. K.; Rutkowski, R.; Kim, P.S. Science 1989,243,538-542. (51) -100([e]₂₂2 + 3000)/33000 was used to calculate the percent u. -helix; Morrisett, J.D.; Jackson, R.L.; Gotto, A.M.,Jr. Biochim. Biophys. Acta 1977, 472, 93-133.

(52) Jasanoff, A.; Fersht, A.R. Biochemistry 1994, 33, 2129-2135.

(53) Tripet, B.; De Crescenzo, G.; Grothe, S.; O'Connor-McCourt, M.; Hodges, RS. J Mol Biol 2002, 323, 345-362. (54) Lahann, J.; Mitragotri, S.; Tran, T.N.; Kaido, H.; Sundaram, J.; Choi, I.S.; Hoffer, S.; Somorjai, G.A.; Langer, R. Science 2003, 299, 311-314.

(55) Brockman, J. M.; Nelson, B. P.; Corn, R. M. Annu. Rev. Phys. Chem. 2000,51,41-63.

(56) Finucane, M.D.; Tuna, M.; Lees, J.H.; Woolfson, D.N. Biochemistry 1999 38,11604-1 1612.

(57) Demers, L. M.; Mirkin, C. A.; Mucic, R. C; Reynolds, R. A., Ill; Letsinger, R. L.; Elghanian.; R. Viswanadham, G. Anal. Chem. 2000, 72, 5535-5541.

(58) Niemeyer, C. M. Trends Biotechnol. 2002, 20,395-401.

Claims

1. A self-assembling peptide linker comprising a peptide belt and a plurality of peptide braces, the peptide belt comprising a plurality of peptide monomer units arranged as a strand, and each of the peptide braces comprising a plurality of peptide monomer units arranged as a strand, wherein the plurality of monomer units of the peptide braces are aligned with the plurality of monomer units in the peptide belt to form a substantially blunt ended coiled coil structure.

2. A peptide linker according to claim 1 in which the peptide belt and the peptide braces together form a leucine zipper.

3. A peptide linker according to any preceding claim in which the peptide belt comprises at least four monomer units.

4. A peptide linker according to any preceding claim in which the peptide belt comprises six monomer units.

5. A peptide linker according to any one of the preceding claims, wherein the sum of the length of the peptide braces is substantially equal to the length of the peptide belt.

6. A peptide linker according to any one of the preceding claims, wherein the peptide monomer units comprise a heptad repeat motif (abcdefg) and/or a hendecad repeat motif (abcdefghijk).

1. A peptide linker according to any one of claims 1 to 5 in which monomer units comprise a heptad repeat motif (abcdefg).

8. A peptide linker according to claim 6 or claim 7, having isoleucine or asparagine at position a.

9. A peptide linker according to claim 8 wherein one monomer unit of at least one of the peptide braces has an asparagine at position a and the peptide belt has an asparagine residue at the corresponding a position ensuring that the peptides braces are aligned with the peptide belt to form a substantially blunt ended coiled coil structure.

10. A peptide linker according to claim 9, wherein the isoleucines are at the remaining a positions.

11. A peptide linker according to any one of claims 6 to 10 having leucine at position d.

12. A peptide linker according to any one of claims 6 to 11 in which the peptide belt has glutamic acid at position e.

13. A peptide linker according to any one of claims 6 to 12 in which the peptide belt has glutamic acid at position g.

14. A peptide linker according to any one of claims 6 to 13 in which a basic amino acid is at position e in the peptide braces.

15. A peptide linker according to claim 14 wherein the basic amino acid is lysine.

16. A peptide linker according to any one of claims 6 to 15 in which a basic amino acid is at position g in the peptide braces.

17. A peptide linker according to claim 16 wherein the basic amino acid is lysine.

18. A peptide linker according to any one of claims 6 to 17 having alanine, glutamic acid or lysine at position b.

19. A peptide linker according to any one of claims 6 to 18 having alanine, glutamic acid or lysine at position c.

20. A peptide linker according to any one of claims 6 to 19 having alanine, glutamic acid or lysine or a tyrosine chromophore at position/

21. A peptide linker according to claim 20 in which at least one monomer unit of the peptide belt and each peptide brace has a tyrosine chromophore at position/

22. A peptide linker according to any preceding claim wherein, at a neutral pH the belt is negatively charged, and the braces are positively charged.

23. A peptide linker according to any preceding claim wherein there is a first and a second peptide brace.

24. A peptide linker according to claim 23 in which a CysGlyGly sequence is added to the N terminus of the first peptide brace.

25. A peptide linker according to claim 23 in which a GlyGlyCys sequence is added to the C terminus of the second peptide brace.

26. A peptide linker according to any one of claims 23 to 25 in which all termini except the C terminus of the first peptide brace and the N terminus of the second peptide brace are capped.

27. A peptide linker according to any one of claims 23 to 26 in which the peptide belt has the sequence: IAALEKEIAALEQEIAALEKEIAALEYENAALEKEIAALEQE

28. A peptide linker according to any of claims 23 to 27 in which sequence of the first peptide brace is: IAALKQKIAALKQKI-AALKYK

29. A peptide linker according to any of claims 23 to 28 in which sequence of the second peptide brace is: IAALKQKNAALKQKIAALKYK

30. A peptide linker according to any one of the preceding claims wherein a first component is attached to the peptide belt and a second component is attached to one of the peptide braces.

31. A peptide linker according to claim 30, wherein one or more further components are attached to one or more different peptide braces.

32. A peptide linker according to any one of the preceding claims wherein a first component is attached to one peptide brace and a second component is attached to a different peptide brace.

33. The peptide linker according to any one of claims 30 to 32, wherein the components are selected from proteins, carbohydrates, solid supports and small inorganic/organic molecules.

34. A method of producing a peptide linker according to any one of claims 1 to 33 comprising providing a peptide belt and a plurality of peptide braces and allowing the peptide belt and the peptide braces to form a substantially blunt ended coiled coil peptide linker.

35. A method according to claim 34 in which the belts and braces are made synthetically and/or recombinantly.

36. A method according to claim 34 or 35 in which the peptide linker is derivatised.

37. A kit for making a peptide linker according to any one of claims 1 to 33 comprising a peptide belt and a plurality of peptide braces wherein the peptide belt and the peptide braces associate to form a substantially blunt ended coiled coil peptide linker.

38. A kit according to claim 37, wherein the belt or one of the peptide braces is immobilised on a solid support.

39. Use of the peptide linker according to any one of claims 1 to 33 to link components together.

40. A use according to claim 39 wherein the components to be linked are attached to separate parts of the peptide linker, namely the peptide belt and the plurality of peptide braces.

41. A method of linking a first component to a second component comprising using a peptide linker according to any of claims 1 to 33 wherein the first component is attached to the peptide belt or a first peptide brace, and the second component is attached to a different peptide brace, wherein on forming the peptide linker, the first and second components are linker together.

42. A method according to claim 41, wherein the first component is a solid support and the second component is a protein.

43. A method of spacing components comprising linking the components using the peptide linker according to any one of claims 1 to 33.

44. A scaffold formed by a plurality of peptide linkers according to any one of claims 1 to 33 linked together via multivalent components.