This article was published as part of the

Peptide- and protein-based materials themed issue

Guest editors Rein Ulijn and Dek Woolfson

Please take a look at the issue 9 2010 table of contents to access other reviews in this themed issue

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online / Journal Homepage / Table of Contents for this issue

3464 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

More than just bare scaffolds: towards multi-component and decorated

fibrous biomaterialsw

Derek N. Woolfson*ab and Zahra N. Mahmouda

Received 4th July 2010

DOI: 10.1039/c0cs00032a

We are entering a new phase in biomaterials research in which rational design is being used to

produce functionalised materials tailored to specific applications. As is evident from this Themed

Issue, there are now a number of distinct types of designed, self-assembling, fibrous biomaterials.

Many of these are ripe for development and application for example as scaffolds for 3D cell

culture and tissue engineering, and in templating inorganic materials. Whilst a number of groups

are making headway towards such applications, there is a general challenge to translate a wealth

of excellent basic research into materials with a genuine future in real-life applications.

Amongst other contemporary aspects of this evolving research area, a key issue is that of

decorating or functionalising what are mostly bare scaffolds. There are a number of hurdles to

overcome to achieve effective and controlled labelling of the scaffolds, for instance: maintaining

biocompatibility, i.e., by minimising covalent chemistry, or using milder bioconjugation methods;

attaining specified levels of decoration, and, in particular, high and stoichiometric labelling;

introducing orthogonality, such that two or more functions can be appended to the same scaffold;

and, in relevant cases, maintaining the possibility for recombinant peptide/protein production.

In this critical review, we present an overview of the different approaches to tackling these

challenges largely for self-assembled, peptide-based fibrous systems. We review the field as it

stands by placing work within general routes to fibre functionalisation; give worked examples on

our own specific system, the SAFs; and explore the potential for future developments in the area.

Our feeling is that by tackling the challenges of designing multi-component and functional

biomaterials, as a community we stand to learn a great deal about self-assembling biomolecular

systems more broadly, as well as, hopefully, delivering new materials that will be truly useful in

biotechnology and biomedical applications (107 references).

1. Introduction

1.1 Rationale for and objectives of this review

This review does not focus on any one class of biomaterial

per se—there are excellent articles on these throughout this

a School of Chemistry, University of Bristol, Cantock’s Close, Bristol,UK BS8 1TS. E-mail: D.N.Woolfson@bristol.ac.uk

bDepartment of Biochemistry, University of Bristol, Medical School,University Walk, Bristol, UK BS8 1TD

w Part of the peptide- and protein-based materials themed issue.

Derek N. Woolfson

Dek Woolfson received hisPhD from the University ofCambridge, UK, working withProf Dudley Williams FRSand Dr Phil Evans onproblems in peptide andprotein folding. He did shortpost-doctoral stints with ProfJanet Thornton FRS(UC London), and Prof TomAlber (UC Berkeley), wherehe honed skills in bio-informatics and biophysics tostudy sequence-to-structurerelationships in proteins. Forthe past 15 years his indepen-

dent group at the University of Sussex, and, since 2005, at theUniversity of Bristol, has focussed on protein design and itsapplication to bionanotechnology and synthetic biology. Hisgroup takes a multidisciplinary approach to its work.

Zahra N. Mahmoud

Zahra Mahmoud received herPhD from the University ofHull, where she was alsoinvolved in the develop-ment of a second-generationtracheo-oesophageal fistulaspeech valve. She currentlyworks as a research associatewith Professor Dek Woolfsonat the University of Bristol;her research interests includedesign and functionalisationof self-assembling coiled-coilbiomaterials.

CRITICAL REVIEW www.rsc.org/csr | Chemical Society Reviews

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3465

Themed Issue—though the ideas and concepts introduced will

be illustrated by reference to specific examples. Rather, our

aim is to layout, and to begin to address, what we see as a

major challenge in biomaterials research today. Of course,

there are multiple challenges in the general area: namely, the

further introduction of truly rational approaches to the design

of such materials; understanding, controlling and exploiting

their dynamics; and the development of methods for

producing biomaterials on large enough scales to be useful

in real-life applications. However, the problem that we focus

on here is the reliable addition of function to what are largely

at present structural materials, or bare scaffolds.

This is challenging for a number of reasons: first, as

supramolecular constructs, we have scant understanding of

the atomic structures of many of these materials, which

frustrates the processes of engineering and rational redesign;

second, and what might seem contradictory to the first point,

many of the scaffolds are so well folded, ordered and even

partly crystalline that they effectively prohibit the inclusion of

the altered or atypical components that might be used to carry

function; thirdly, many of the materials are non-covalent

assemblies and attempts to modify them post-assembly fail

because of the underlying fragility of the systems, or

the forcing conditions needed for some bioconjugations.

Nonetheless, researchers are beginning to succeed in

assembling multi-component systems, or decorating pre-

assembled materials, giving considerable hope to addressing

the challenge we have laid out.

1.2 Types of material and how these impinge on approaches to

functionalisation

Mainly, we consider three types of peptide and protein fibrous

materials here: (a) those based on amyloid-like structures,

which can be formed from natural proteins, naturally derived

peptides, or peptides of de novo design; (b) those based on

a-helical assemblies, which usually employ peptides of de novo

design; and (c) engineered peptide-synthetic hybrids, in

particular peptide amphiphiles (PAs). This is primarily because

these self-assembling systems have been developed sufficiently

to allow their decoration and functionalisation. Other structures

such as designed collagen-like assemblies and silks will be

introduced as appropriate in later sections of this review.

(a) Amyloid-like structures. The term amyloid has a specific

meaning: it refers to extracellular, insoluble, fibrous, protein

aggregates associated with diseases such as Alzheimer’s.

However, as many peptides and proteins—both natural and

engineered, and implicated in disease or not—can be induced

to form similar assemblies, including similar underlying

secondary and quaternary structures,1 the more-general term

‘‘amyloid-like structure’’ is often preferred. In this generic

structure, small model peptides, fragments of natural proteins,

or even whole proteins, undergo conformational switches

to largely b-structured states in which multiple b-strandshydrogen bond to form extended sheets, these pair to form

the generic cross-b-structured fibrils. These may or may not

bundle to form thicker fibres. In these structures, the b-strandsrun perpendicular to the long axis of the fibrils and fibres,

Fig. 1a.1

Sequence-to-structure relationships in peptides that form

amyloid-like structures are emerging, but the field is still in its

infancy in this respect. This is somewhat different from the

a-helical assemblies described next, where such relationships

tend to be better defined partly because of the better under-

standing that we have for helical assemblies from work

on free-standing and coiled-coil a-helical systems.4 None-

theless, in terms of rational peptide design, rules of thumb

for engineering peptides that form amyloid-like structures

include: using a high proportion of b-structure-favouringresidues; aromatic side chains; and employing sequence

patterns in which hydrophobic (H) and polar (P) residues

alternate, (HP)n.

(b) a-Helical assemblies. In amyloid-like structures, inter-

chain backbone hydrogen bonding plays a significant role

in the self-assembly process. For a-helical supramolecular

assemblies this is not, indeed cannot, be the case. This is

because the majority of the backbone hydrogen bonding of

the a-helix is tied up locally within the helix. This has

advantages and disadvantages for design: one advantage is

that potentially promiscuous backbone-backbone interactions

are avoided, or at least reduced, which increases the level

of control possible compared with that for amyloid-like

assemblies; a disadvantage might be stated in the question,

then how are inter-chain interactions and higher-order assemblies

achieved? The answer to this comes in part from clever protein

designs, and partly through serendipity. The design aspect is

that sequence-to-structure relationships for natural helix–helix

assemblies are used, and slightly altered to drive new modes of

helix–helix interaction that promote supramolecular assembly;

essentially, a-helical peptides can be designed as building

blocks, or tectons, for assemblies comprising many millions

of organised peptides.5 The serendipity is that, when reduced

to practice, rather than making long thin fibrils, most of these

designs form long thick fibres. It turns out that this makes

them more interesting, robust and potentially more useful than

initially envisaged.

Several groups have now succeeded in making a-helix-based fibrous systems,6–11 which have been reviewed by us

elsewhere.12,13 However, the necessary details and differences

between these systems are described in section 4 of this review.

These designs rely on programming polypeptide sequences to

form amphipathic helices that associate—usually as dimers,

trimers and pentamers. As such, the designs tend to be based

on the well-understood a-helical coiled-coil motif,4 and have

underlying heptad repeats, (HPPHPPP)n. However, with a

couple of notable exceptions,10,11 unlike in natural coiled-coil

assemblies, where the bundles are discrete or blunt ended, in

the fibrous designs the helices are programmed to pack in

slipped arrangements. This leaves overhangs, or sticky ends,

which, in turn, assemble end-to-end to drive fibrillogenesis.

This concept is illustrated in Fig. 1b.

(c) Peptide amphiphiles (PAs). Hartgerink and Stupp’s

concept for the peptide amphiphiles (PAs),3 presents a very

different approach to engineering soft materials: here

self-assembling and functional units are combined in one

molecule from the outset. In the original PA design,3 a

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3466 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

16-carbon alkyl chain is conjugated to the N-terminus of a

polar ter-functional linear peptide with the sequence,

Cys4-Gly3-SerP-Arg-Gly-Asp, Fig. 1c. The alkyl chain, and

the overall amphipathicity of these molecules drives assembly

into rod-like, fibrous micelles, in which the alkyl chains are

buried and the C-termini of the peptides are exposed; these

structures are stabilised by a corona of interchain disulfide

bridges between the cysteine block; and the four C-terminal

amino acids provide water solubility and function. Stupp’s

team has confirmed these concepts and basic aspects of

the design and assembly through a series of biophysics,

microscopy and rheology studies over the past decade.14,15

Each of the general classes of material a–c has advantages

and drawbacks for decoration, which can limit or even dictate

the optimal route(s) taken to functionalising the systems. For

instance, the amyloid-like materials tend to have one or both

of their N- or C-termini free. Thus, the functional moieties, be

they small molecules, other peptides or whole peptides, can

simply be appended to the amyloidogenic polypeptides prior

to assembly. Similarly, the design principles for the PAs

specifically considers the need to incorporate functional motifs

with the molecular design. Clearly in both cases a and c, this

carries many advantages, but the possible lack of control over

assembly can be an issue; often these systems employ a single

self-assembly unit, and are subject to spontaneous assembly,

or thwarted by nucleation kinetics. As outlined in section 4,

the decoration of at least some of the a-helical systems is less

straightforward; as the termini are not necessarily free for

conjugation for example. Nonetheless, the experience and

knowledge that the community has in design and engineering

a-helical systems when fully brought to bear in this area of

materials could have a major impact.

2. Decoration of self-assembled, fibrous

biomaterials

Broadly, there are two general approaches to making multi-

component and decorated materials: top-down assembly,

using a variety of patterning techniques; and bottom-up

assembly, exploiting controlled self-assembly of monomeric

units. Top-down processes are used extensively for the

fabrication of photoptical devices, nanowires and so on. More

recently, this approach has been used to pattern surfaces for

biological applications.16 However, the approach presents

limits for incorporating features at the nanoscale. This review

focuses on routes in the bottom-up design of both straight-

forward and complex (multi)functional biomaterials via the

self-assembly of peptide, and peptide-containing systems.

Several routes have been taken here, which can be divided as

follows. Firstly, there is the construction of monomeric units

that contain both the instructions for self-assembly, and a

functional moiety; we refer to these as Route 1, or co-assembly

approaches. Secondly, in Route 2, or post-assembly approaches,

bare scaffolds are assembled from more-basic building blocks,

and then decorated either through covalent (Route 2a), or

non-covalent modification (Route 2b). Thirdly, in Route 3, or

templating approaches, pre-assembled fibrous materials are

used as templates to draw down the assembly or condensation

of inorganic or polymeric materials. These routes are

illustrated in Fig. 2.

Fig. 1 Types of peptide-based self-assembling fibrous biomaterials. (a) Amyloid-like assemblies. Small peptides form b-strands that hydrogen bond

into b-sheets and then assemble to form cross-b-structured fibrils. Reproduced in part from ref. 2, by permission of the National Academy of

Sciences, copyright r 2002. (b) a-Helical assemblies. Amphipathic a-helices can be designed to interact offset and form extended coiled-coil fibrils

that bundle into long fibres. (c) Peptide amphiphiles. Here the information needed for both self-assembly and functionalisation are encoded within

the same hybrid molecule. Reproduced in part from ref. 3, with permission of Science copyright r 2001. (d) Collagen-like assemblies. Stretches of

ProHypGly sequence form polyproline-type helices that associate into right-handed triple helices. These bundle further to form collagen-like fibres.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3467

We find these classifications useful, but we recognise that

they are broad and cover a wide range of different materials,

chemistry and work from many groups. As a result, they can

be split further. For instance, in the co-assembly approach for

example, building blocks for assembly and function can be

combined in the initial pre-assembled construct, as in

Stupp’s peptide amphiphiles,15 or Barker’s amyloid-forming

protein-cytochrome fusions.18 In the non-covalent approach,

functionalised building blocks can be incorporated into the

fabric of the assembling material if it is permissive, as

exemplified by Yu’s work on collagen functionalisation;63 or

through the design or selection of new binding motifs as we

demonstrate for our own Self-Assembled peptide Fibres

(SAFs).19 The final approach of templating is somewhat of a

special case, which we split according to the types of materials

that are deposited on the scaffold.

Finally, we give a short history cataloguing our attempts to

decorate the SAFs, which have met with different levels of

success; and following this we describe the work of others in

this area who work on similar systems comprising a-helicalbuilding blocks.

2.1 Route 1: co-assembly of covalently linked structural and

functional moieties

The first approach to generating peptide- or protein-based

fibres that display a functional moiety is simply to construct

a single molecule containing both the self-assembling and

the functional components in question. There are three

ways in which this might be envisaged: (1a) covalent attach-

ment of the moiety to polypeptide during peptide synthesis;

(1b) if the functional moiety is itself a peptide or a

protein, gene fusions can be designed and expressed recombi-

nantly; and (1c) hybrid peptide systems in which the peptide

parts act as the self-assembling unit, the functional moiety,

or both.

In all of these cases, functionalisation is covalent and

performed prior to assembly. Of course, this mitigates some

problems associated with incompatibility of bioconjugation

chemistry and self-assembly. However, it places restrictions on

the systems too: namely, and obviously, that the resulting

constructs must remain competent for self-assembly; and,

conversely, that the self-assembly process must not preclude

or otherwise hamper the accessibility or activity of the

attached functionality. For self-assembling peptides and

proteins, this usually means having an available N- or

C-terminus free for covalent chemistry; though, side-chain

modifications are also possible. Indeed, the latter are becoming

more-widely used with wider access to modern peptide chemistry,

and the possibility of incorporating non-proteinogenic amino

acids into proteins via new recombinant methods.

Route 1a: peptide-small molecule constructs. Modern solid-

phase peptide synthesis (SPPS) is opening a wide range of

possibilities for bringing function to synthetic peptides and

even small proteins.20 Peptides of up to B50 amino acids can

now be made routinely by SPPS. In these methods, the peptide

chain is made by a series of catalysed condensation reactions

between a resin-bound polypeptide with a free N-terminal

amino group, and an amino acid in solution with a protected

N-terminus and an activated C-terminal carboxylic acid.

During these steps, the side chains of the growing polypeptide

chain and the incoming amino acid are protected with

chemistry that is orthogonal to that used to free the

N-terminus and effect the condensation reactions. In this

way, the peptide is synthesised from its C- to its N-terminus,

and, at the end of the process, a complete, side-chain-protected

peptide is attached to the solid support. At this stage, the

N-terminus can be liberated separately—as can certain

side-chain protecting groups—to allow a variety of functional

groups to be added through straightforward acid-amine

conjugation or alternative chemistries. Finally, the (remaining)

side-chain protection can be removed, and the peptide cleaved

from the resin simultaneously to produce the free peptide

ready for self-assembly.

There are numerous examples of this chemistry being

applied to add fluorophores, biotin (for later capture by

streptavidins), small peptides and similar moieties to the

N-termini of peptides, but here we focus on functionalising

self-assembling peptides. This particular area has been

reviewed by Channon and MacPhee recently,21 but we offer

Fig. 2 Routes to decorating self-assembling systems. (a) Co-assembly

of covalently linked structural and functional moieties. (b) A func-

tional protein can be (genetically) fused to a self-assembling peptide/

protein. (c) Post-assembly covalent decoration of scaffolds using

techniques such as click chemistry. (d) Structural mimics or (e) non-

covalent interactions can be exploited for functionalisation of

pre-formed scaffolds. Reproduced in part from ref. 17 with permission

of Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, copyright r

2010. Peptides/protein fibres can be used to template hollow tubes (f),

and conducting nanowires (g).

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3468 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

up a few illustrative examples from the primary literature.

Most of the examples centre on using amyloid-like assemblies.

For example, Channon et al. describe the assembly of

modified peptides based on the 105–115 fragment from

transthyretin (referred to as TTR), which forms amyloid-like

assemblies. In this case, they add a fluorophore to the

N-terminus of the peptide to give F-TTR. Assembly is driven

solely by the peptide sequence, with the fluorophore still

retaining its functionality, but not contributing to the

assembly process; i.e., F-TTR assembles to 7 nm-wide fibrils

as observed by transmission electron microscopy (TEM),

which show meridional and equatorial reflections at 4.7 and

10 A, respectively, in X-ray diffraction indicative of cross-bstructure.22 The fluorophores in these assemblies exhibit new

chiro-optical properties, which, the authors suggest, may find

use or provide inspiration in the development of opto-

electronic materials for photovoltaics or synthetic photo-

synthesis. Indeed, towards the second goal, the team has

recently demonstrated the co-assembly of two independent

luminescent moieties.23 In these cases the ‘‘conjugations’’ are

straightforward, because the fluorophores are added to the

N-terminus after peptide synthesis; indeed, in the first

example, it is simply the Fmoc protecting group retained from

peptide synthesis.

N.b. The development of shorter, Fmoc and similarly

N-terminally conjugated peptides, is a rapidly growing sub-

field being explored and reviewed well by Gazit, Xu, Ulijn,

Adams, Zhang and others.24,25 Though we note in these cases

the N-terminal aromatics and fluorophores usually do play a

significant role in the assembly of these smaller peptides

compared with the TTR-based systems.

Also somewhat related to the example from Channon et al.,

though strictly it comes under Route 2b in our categorisation,

Welland and co-workers describe how concentrated amyloid-

like fibres can be cast into ordered gels, that bind and orient

fluorophores non-covalently.26 One advantage here is the use

of relatively cheap, commercially available starting materials:

the protein used is hen-egg lysozyme, and the fluorophore is

Thioflavin T.

Extending this type of study in yet another direction, Gras

et al. describe the construction of bio-active peptides based on

TTR.27 In this case, the authors make C-terminal extensions

with a Gly-Gly spacer followed by the integrin-binding peptide

Arg-Gly-Asp (RGD). The resulting peptides still form

amyloid-like assemblies that display the expected dye-binding

tinctorial properties, fibre morphology by electron micro-

scopy, and cross-b-structure in X-ray diffraction. Further-

more, dansyl fluorophore appendages, also at the C-termini,

are used to demonstrate the exposure and availability of the

C-terminus, and that blends of the parent TTR1 and

TTR1-RGD can be made; both are confirmed by immunogold

labelling using an anti-dansyl fluorophore antibody and

visualisation by electron microscopy. The RGD-decorated

fibres are used in cell studies with fibroblasts to demonstrate

their use as substrates to support cell adhesion in tissue culture.

Gras’s work adds to a growing body of studies on designed

and natural peptides that form amyloid-like assemblies, and

particularly hydrogels, being applied in the burgeoning area of

3D cell culture and tissue engineering. On the design front for

example, research groups like those of Zhang, Pochan &

Schneider, and Aggeli are contributing here: Zhang’s group

has added RGD-based peptides to his model amyloid-like

gel-forming system, RADA16;28,29 though the MAX series

of peptides from the Pochan and Schneider groups have not

been engineered in this way, they do support 3D cell growth,30

and interestingly show antimicrobial activity;31 and Aggeli and

co-workers have considered the requirements for expanding

the field to real-life applications, including the need to move to

recombinant production in place of peptide synthesis.32,33

Regarding other naturally derived peptides—in particular

those not from proteins normally associated with amylo-

idogenic peptides or amyloidoses—Ohga et al. demonstrate

that certain proteolytic fragments of laminin (a glycoprotein

component of basal lamina) form amyloid-like fibres.34,35

These can be extended with N-terminal RGD-containing

sequences, and retain the ability to form amyloid-like

structures. The resulting functionalised materials support cell

growth, and neurite-like outgrowth from a model rat phaeo-

chromocytoma (PC12) cell line in culture. The second

property is linked to the sequence motif Ile-Lys-Val-Ala-Val

(IKVAV), which some of the peptides also contain making

these assemblies multifunctional.

Route 1b: recombinant peptide-protein fusions. Recombinant

DNA technology and, with it, the ability to clone, mutate and

fuse genes for proteins, and then express these in tractable host

cells is well established.33 Nonetheless, the increased efficiency

and reduced cost of making synthetic DNA, together with the

non-specialist kit or standardised approach to cloning and

expressing such pieces, are influencing the development of the

new field of synthetic biology. In essence, the aim here is to

design de novo, or to engineer existing biological components,

and to put these together in ways to make new and interesting

systems and devices, which, hopefully, will have some new and

useful properties.36 In these respects, the design and engineering

of functional peptide- and protein-based materials, parti-

cularly when translated into recombinant production, can

be considered a legitimate aspect of synthetic biology.5

However, it is one thing to sketch out ideas for basic bio-

molecular components, it is another to start fusing them

together in predictable, and useful ways. Nonetheless, several

studies offer hope and inspiration here.

For example, using the yeast protein Ure2p, Baxa et al.

demonstrate that functional proteins can be piggybacked onto

a self-assembly framework to give functional fibres.37 The

N-terminal ‘‘prion’’ domain—either residues 1–65 or 1–80,

depending on the construct used—of Ure2p protein is natively

unfolded, but can assemble to form b-structured amyloid-like

fibrils. Baxa and co-workers fused these regions to four

proteins—three enzymes, barnase, carbonic anhydrase,

glutathione S-transferase, and the green fluorescent protein.

In all cases the soluble fusion proteins show unfolded Ure2p

domains, but assemble to b-structured fibrils. Moreover, and

though to different degrees, the appended natural proteins are

all active. Similarly, Barker et al. have fused the gene for

cytochrome b562 to the amyloidogenic SH3 dimer, which is

capable of independently forming fibrils.18 Addition of

cytochrome b562 does not interfere with fibril formation, and

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3469

cytochrome b562 when displayed on the fibre retains functionality

by binding metalloporphyrins, though only half of the

potential binding sites are occupied.

An area where recombinant production of materials has a

distinct advantage is for large, repetitive and otherwise

intractable fibrous proteins such as silks.38,39 The rules relating

to sequence-to-structure in these systems are only just being

unravelled. Nonetheless, they have some superb materials

properties ripe for exploitation as engineered biomaterials.

The sequences are modular and this has been exploited, by the

Kaplan and Scheibel groups for example, to produce and

express synthetic genes. In terms of decoration, Kaplan’s

group in particular has shown that these can then be used to

introduce single and multiple cell-adhesion motifs near the

termini of fibres to generate bioactive tissue-engineering

scaffolds.40,41 In this general area, others are turning their

attention to other elastomers such as arthropod resilins,42

though at this stage this work has focused on producing

recombinant model proteins, and to our knowledge, these

have yet to be decorated.

As noted by Kyle and co-authors,33 the move towards

recombinant production of self-assembled systems, either as

independent moieties or functionalised as above, will become

increasingly important if the field is to advance and fulfil

expectations towards real applications. Though it is not with-

out difficulties—amongst these is the overproduction of small,

often ephemeral peptides; and the possible formation of

cytotoxic oligomers or matured fibres—there is encourage-

ment that the community is headed in this direction.43

Route 1c: hybrid peptide-containing systems. Perhaps the

classic examples of materials in which the components for

self-assembly and function are directly and deliberately

incorporated within one molecule are Stupp’s peptide

amphiphiles (PAs), Fig. 3.3 In the original PA, the polar part

of the incorporated peptide sequence, -SerP-Arg-Gly-Asp, has

a dual function: the phosphoserine is for hydroxyapatite

templating; and, as mentioned above, the RGD sequence is

for cell binding. Stupp’s team has increasingly demonstrated

potential biomedical applications for the PAs.15 For example,

different PA fibres have been used as tunable scaffolds for

tissue engineering,44,45 including the capture and controlled

release of growth factors,46 and as bioactive surfaces for bone

implants.47

In a different approach to multi-domain amphiphiles

Hartgerink and colleagues make ABA-type systems from

purely peptidic blocks.50 Here, the amphiphilic B block drives

self-assembly into nanofibres around 6 nm in diameter,

whereas the hydrophilic A block is composed of charged

amino acids and determines the conditions under which

assembly takes place. Moreover, functional motifs for

cell-adhesion and enzyme cleavage can be placed in the central

B block to create functional responsive bioactive gels.51

2.2 Route 2: post-assembly decoration

Functionalisation performed after assembly of the scaffold

poses different challenges to the assembly of combined

structural and functional components. Moreover, the new

functions can be added covalently (Route 2a) through bio-

conjugation, or non-covalently (Route 2b) through binding. To

make such post-assembly modifications requires accessible

modifiable groups on the surface of the scaffold, which is

pertinent to both Routes 2a and 2b; and the scaffold to be

tolerant of the conditions used for functionalisation, which is

usually more of an issue for covalent modification. Thus, there

are potential restrictions and drawbacks of taking the post-

assembly route. However, advantages of separating assembly

and functionalisation in this way include the possibility of

making general, and potentially more-versatile materials.

These could then be applied in a variety of different

applications without recasting the underlying self-assembled

scaffold. i.e. For a given scaffold, and a robust, general way of

appending cargo to it post-assembly, a wide variety of new

functional biomaterials can be envisaged.

Fig. 3 Functionalisation of peptide amphiphiles. (a) Schematic of a bioactive peptide amphiphile sequence displaying an IKVAV motif (adapted

from ref. 15). Peptide amphiphile nanofibres (b), and resulting implants used in bone regeneration (c). Reproduced in part from ref. 48, with

permission of Elsevier Ltd., copyright r 2010 . (d) Heparin-binding amphiphile nanofibres can support angiogenesis. Reproduced in part from

ref. 49, with permission of Elsevier Ltd., copyright r 2009 .

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3470 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

Route 2a: post-assembly covalent modification. There are few

examples in this subtopic because of the tension between the

conditions needed to make covalent modifications, and those

required to maintain the self-assembled materials. However,

we suspect that click chemistry will transform this particular

area, facilitating new routes to functionalised self-assembled

peptide and proteins materials. Indeed, a search of Web

of Science combining the terms ‘‘click chemistry’’ and

‘‘self-assembly’’ and protein/peptide returned less than 20

papers, most of which were from the last year (2009–2010).

Thus, it seems to be an emerging and promising area.

However, most of the papers we found would be best placed

in the previous section, as they involve the conjugation of

peptides and, in one case proteins,52 to polymers to make

diblock and triblock systems that subsequently self-

assemble.53–55 Though a fascinating area for research, this

field of peptide-polymer hybrids is broad with a large literature

that is beyond the scope of this review. Therefore, we refer the

reader to a number of excellent recent reviews, which expand

this and the previous sections generally.56–58

We examine the area of click chemistry as applied to purely

peptide-based self-assembled systems with reference to our

own work in the area in section 4. Click chemistry

encompasses several types of reactions, of which the 1,3-dipolar

Huisgen cycloaddition is the most popular because of ease of

synthesis of the intermediates, high orthogonality, relatively

fast reactions kinetics and good yields. However, copper-based

click chemistry has not been widely explored for functionalisa-

tion of biomaterials because of concerns over the toxicity of

copper. However, we and others have shown that copper

can be successfully removed from gels after functionalisation;

additionally copper-free click is increasingly becoming accessible.59

As an example for this section one recent paper is parti-

cularly noteworthy even if a little off the main topic of the

review. DeForest et al. describe an elegant modular hydrogel

system that supports cell growth, Fig. 4.60 The components

comprise: a ‘‘hub’’ with four poly(ethylene glycol) (PEG) arms

each terminated by an azide functional group; a synthetic

peptide ‘‘linker’’ incorporating multiple features including:

cyclooctyne-containing terminal residues, a central alkene-

containing side chain, and a protease cleavage site; and

thiol-containing peptides that harbour fluorophores or other

protease sites to provide ‘‘functional appendages’’. Two ortho-

gonal click reactions are employed in the assembly process:

firstly, a copper-free variant of a 1,3-dipolar Huisgen cyclo-

addition between azides and alkynes is used to bring together

hubs and linkers to produce a highly cross-linked material,

which forms a physical hydrogel in aqueous buffer. Secondly,

photocatalysed thiol-ene reactions are used to tag the alkene-

containing hydrogel framework with the thiol-containing

functionalised peptides. Impressively, this can be done in a

top-down manner to allow cytocompatible photopatterning

with tens of micron resolution. In this way, the authors

prepare a number of hydrogels that can be visualised,

patterned, used to support cell growth and enzymatically

degraded.

Route 2b: post-assembly non-covalent binding. This route is

attractive for the same reasons that biology employs weak

non-covalent interactions—usually combinations of hydrogen

bonds, van der Waals’ forces, and salt bridges—to assemble

functional complexes: essentially, such forces allow for

control, specificity and reversibility; and they present possibi-

lities for building up complex (multi-component) systems. In

biomaterials design, they offer another potential advantage of

mild, biocompatible conditions for decoration. Balanced

against this, it is often difficult to rationally design specific

binders for a substrate (even if its surface structure is known);

and any binding will be reversible to some extent, and the

dynamics of cargo binding, release and, therefore, longevity on

the substrate may present problems.

Non-covalent decoration via hydrophobic interactions.

Peptides that form well-ordered supramolecular structures

are good candidates for non-covalent decoration, and make

even the notion of rationally designing binding peptides

Fig. 4 Functionalisation of polymer networks using copper-free click chemistry. A polyethylene glycol hub with terminal azides (a) is cross-linked to

cyclooctyne-bearing functional groups (b) to produce a physical hydrogel (c) with regularly arranged functional groups (R2; green). Reproduced in

part from ref. 60, with permission of Macmillan Publishers Limited, copyright r 2009.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3471

possible. For example, Mihara’s group have reported a

peptide design, FI, that self-assembles to form straight, highly

ordered, tubular amyloid-like structures.61,62 Interestingly, FI

first assembles into protofibrils 5–17 nm in diameter, these

align side-by-side to form ribbons, which twist to form straight

fibres 80–130 nm in width andB10 mm in length. This packing

gives striations in negative-stain TEM—indicative of a high

level of order—and it is proposed that it exposes some of the

hydrophobic residues within the fibre. On this basis, the group

has designed anchor molecules to interact with these hydro-

phobic regions non-covalently.17 These consist of biotin linked

to a hydrophobic dodecyl group with or without a spacer unit.

Assembly of F1 is not perturbed by either anchor, and

subsequent incubation with gold-conjugated antibody shows

that the anchors were successfully recruited to the surface of

the fibres. To expand this style of non-covalent anchoring, the

dodecyl component can be replaced with Ile-Ile or Phe-Phe

dipeptide units to similar effect. Moreover, these dipeptide

binders can be extended via b-alanine linkers with imino-

diacetate moieties for Ni-chelation and subsequent capture

of hexahistidine-tagged proteins. Using these constructs, the

group present some striking images of nanoparticles arranged

with regular periodicity along the edges of fibres, Fig. 5.17

Non-covalent decoration using structural mimics. Yu and

co-workers describe how peptides that mimic collagen

sequence and secondary structure, so-called collagen-mimetic

peptides (CMPs), can be incorporated into the triple-helical

structure of collagen during the assembly process and so be

used to bring function to this natural scaffold.63 Naturally

occurring fibrous collagens form right-handed triple helices

due to underlying Pro-Hyp-Gly sequence repeats (where Hyp

is hydroxyproline), Fig. 1d. The CMPs are short, synthetic

peptides made up of such repeats. Yu and colleagues reason

that CMPs should associate with partly denatured collagen,

and demonstrate this as a series of fluorophore-labelled CMPs

with varying melting temperatures (TM): long, for instance

(ProHypGly)10-based CMPs with high TMs interact with

collagen films equilibrated at 80 1C, but not at 25 1C;

shortening the core sequence, to (ProHypGly)6, lowers the

TM to 25 1C, and this modified CMP interacts with collagen

films at 30 1C. Importantly in the second case, the temperature

is lower than the TM of collagen itself (37 1C), indicating that

denaturation of CMP is more critical for the interaction than

denaturation of collagen. To study the binding site of CMPs,

(ProHypGly)7 has been conjugated to gold nanoparticles via a

cysteine residue.64 This shows preferential decoration of type I

collagen fibres in the gap regions, which arise from staggered

packing of the triple-helices along the long fibre axis, and

appear as dark bands every 67 nm in negative-stain TEM. The

group suggest that the gaps correspond to regions low in

hydroxyproline, and that the CMPs supplement this acting

as triple-helix stabilisers and, hence, have increased affinities

for these regions. CMPs may be exploited in tissue engineer-

ing. For example, it is possible to make hybrid PEG-CMP

hydrogels, which retain cell-secreted collagen, thus allowing

the cells to grow in an environment more-closely mimicking

the natural extracellular matrix.65 Further modifications to

these gels using CMPs have been tested, including the addition

of multiple anionic charges to the N-terminus to recruit

VEGF.66 The VEGF-CMP charged complex is released from

the gel at a sustained rate, allowing the development of Hu-

man Umbilical Vein Endothelial Cells (HUVEC) with a

tubular morphology indicative of three-dimensional growth.

Non-covalent decoration using complementary pairing.

Related to Stupp’s earlier PA work, Li and Stupp have

engineered new PAs after Yamada and co-workers.67 In these

Fig. 5 Non-covalent decoration of self-assembling fibres with amyloid-like structure. (a) Bi-Da and Bi-ADa dodecyl anchors bind non-covalently to

F1-peptide fibres via hydrophobic interactions. (b) Replacing the dodecyl chain with Ile-Ile units allows the anchors to bind with a regular

periodicity. Reproduced in part from ref. 17 with permission of Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, copyright r 2010.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3472 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

constructs, fatty acid-like dialkyl-chains and monoalkyl chains

are linked through a tripeptide unit. These wedge-shaped

molecules assemble to nanofibres in organic solvents burying

the monoalkyl moieties and leaving the dialkyl-chains

exposed. In this case, Li and Stupp append thymine to the

terminus of one of the latter chains. This can then be

non-covalently decorated with gold nanoparticles functiona-

lised with diaminopyridine (DAP), which forms comple-

mentary hydrogen bonds with the base, to render long, linear

gold nanowires.

Returning to water-soluble PAs, Guler et al. and Stupp,

have decorated yet another variant of the PA nanofibres

post-assembly using a biotin-avidin label.68 These designs

differ in a number of respects from the original PAs: first,

the alkyl chain, which drives assembly of the rod-like micelles,

is appended to a C-terminal side-chain of a central poly-

peptide; second, at the N-terminus of the polypeptide two

branches are spurred off from lysine side chains. These

branches are RGDS peptides with N-terminal biotins. The

resulting PAs are again wedge-shaped and form nanofibres

that gel. The surface biotin moieties can then be decorated

with BODIPY-NeutrAvidin.

2.3 Route 3: templating hard inorganic materials on soft

biomolecular assemblies

The final approach to decorating self-assembled peptide- and

protein-based biomaterials that we discuss is templating. This

bridges the gap between soft, self-assembled scaffolds, which

afford control over assembly and nanostructure at the 2-D and

3-D levels, and systems that cannot, or do not readily self-

assemble, but bring new functions to the resulting composite

materials. In general terms, this is another broad field,69 which

we cannot do justice to here, so we present a small number of

illustrative examples.

In templating, usually inorganic materials—though there

are examples for organic and polymeric materials—are deposited

on the surfaces of the scaffolding materials. This is either done

passively, which usually relies on recruitment of charges akin

to layer-by-layer deposition,70,71 or more-actively, which often

involves the inclusion of thiols (usually cysteine residues) into

the peptide and protein scaffolds. Materials that have been

combined with biomolecular assemblies in this way include,

gold and silver nanoparticles and metals, clays and silica,

and polyelectrolytes. The challenge is to template these in a

controlled fashion, to form functional conducting wires or

biomimetic substitutes for example.

Route 3a: towards metallic nanowires and related bio-

inorganic structures. Building on amyloid-like assemblies,

Reches and Gazit have exploited the self-assembling properties

of aromatic dipeptides to create silver nanowires with uniform

dimensions, Fig. 6.72 At high concentrations, diphenylalanine

precipitates out of hexafluoro-2-propanol as self-assembled

hollow nanotubes with outer diameters of B100 nm. The

hollow of these tubes can be filled with silver ions, which are

subsequently reduced to silver metal using citric acid as a mild

reducing agent. Moreover, the outer peptidic shell can

be removed with proteinase K to leave free-standing silver

nanowires of B20 nm diameter and almost microns in length.

Thus, the authors have succeeded in passively casting metal

wires using a peptide scaffold.

Similarly, Scheibel et al. use a variant of the N-terminal-to-

middle, ‘‘NM’’, region of yeast prion protein Sup35p, which

assembles into fibres B10 nm wide and up to microns in

length, to template gold nanoparticles.76 In this case, the initial

recruitment is actively encouraged by using a cysteine-

modified variant of NM. Again, metal wires are produced

by reduction of either silver or gold ions from solution. In this

case, the resulting silver and gold nanowires coat the surface of

the fibres, and are B100 nm across. With both metals, the

wires show low-resistance and ohmic metallic properties.

Further on this theme, other early pioneering studies in this

area come from Belcher and co-workers. Along with others,

such as Naik,69 this group show how metal recruitment can be

specified and controlled more tightly using peptide aptamers

selected using phage display against the target metal or

inorganic material.77 In these studies, the bacteriophage M13

is used both as a vehicle for peptide selection, and as a scaffold

for display and decoration, Fig. 7. M13 is a single-stranded

DNA virus widely used in molecular biology. The virus

particles are filaments with dimensions B7 nm in diameter

and B1 mm in length, which display a number of different

proteins on their surfaces in copy numbers ranging from

B3–3000. The genes for these proteins can be modified to

allow peptides and proteins to be displayed on the viral

surfaces; hence the term phage display. Belcher, Naik and

others have used this system to display libraries of peptides

and to select from these those that specifically and tightly bind

inorganic surfaces. The cunning tricks that allow this are:

(1) that binding phage can be enriched and amplified through

rounds of selection and infection into Escherichia coli; and

(2) that the sequences of the selected peptides can be deter-

mined via DNA sequencing as they are barcoded in the viral

DNA. In this way, Belcher’s team, for example, has developed

phage-templated silver,78 and gold and cadmium nanowires.79

Recently, the group has focussed on developing more-

ambitious phage-templated systems as a basis for highly

conductive lithium-ion batteries,80 and porphyrin assemblies

for light-driven water oxidation.81

Other recent examples of note in this general area include:

the production of metal-insulator-metal, trilayered, coaxial

nanocables from Gazit’s group.85 In this case, the afore-

mentioned diphenylalanine tubes that sequester silver are

supplemented by cysteine-containing peptides that bind the

surface of the structures to provide a second, outer template

for metalation. Recently, Ostrov and Gazit have turned to the

prokaryotic tubulin homologue, and filamentous protein Z

(FtsZ), which, in the presence of GTP or calcium, assembles

into nanometre-scale fibres. Through genetic engineering,

various metal-binding peptides with different metal-binding

affinities have been fused to the N-terminus of the protein,

allowing a series of nanowires to be generated.86 Mitraki et al.

used the self-assembling octapeptide peptide Asn-Ser-Gly-Ala-

Ile-Thr-Ile-Gly (NSGAITIG) from the fibre protein as a

template. This forms amyloid-like structures, and N-terminal

cysteine variants can be used to sequester gold, silver and

platinum nanoparticles to impressively high coverage.74 Yet

another structure has been explored as a template by Raines

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3473

and co-workers. They use collagen-like self-assembling peptides

and electroless silver plating (silver enhancement) to generate

nanowires. The basic collagen repeat of Pro-Hyp-Gly is

modified by replacing one of the hydroxyprolines by lysine,

which, in turn, is used to recruit amine-reactive gold nano-

particles, followed by silver enhancement.87

Route 3b: templating minerals on soft materials. As already

covered, the original PAs from Hartgerink and Stupp template

the mineralisation of hydroxyapatite via the phosphoserine

incorporated in the peptide component.3,88 PAs also assemble

into gels when incubated with calcium, which decorates the

outside of the fibres to form nucleation points for hydroxy-

apatite mineralization. This system has been expanded further

to template hydroxyapatite crystals in 3-D using the enzyme

alkaline phosphatase.89 Additional rigidity and stability can be

added to the system by filling these amphiphile gels into

titanium foam scaffolds.47

Similarly, Hartgerink et al. have used PAs bearing an RGD

motif and a enzyme-cleavable sequence to encapsulate two

types of dental stem cells.90 When grown in the presence of

osteogenic supplements, these PAs support growth of both

soft cells and osteoblastic cells, which mineralize the matrix,

making this a useful scaffold for dental tissue regeneration.

On a related theme, Aggeli and co-workers have used their

amyloid-like b-tapes as a scaffold to promote hydroxyapatite

crystallization and tooth mineralization.91

4a. Anecdotes from studies towards multi-component and

decorated SAFs

Over the last decade, we have explored the design, assembly

and application of a Self-Assembling peptide-based Fibre

(SAF) system.7,92–94 Unlike the majority of the systems

described above, which tend to employ a single self-assembling

molecule, the basic SAFs comprise two peptides that only

assembly when mixed. The peptides are of completely de novo

design, based on well-established sequence-to-structure

relationships for a-helical coiled coils.4 A key difference

between the SAFs and natural, or other designed coiled coils

is that they are designed to form offset, or staggered hetero-

dimers.7 These so-called sticky ended dimers are the building

blocks for fibrillogenesis. We have performed a detailed

analysis of the folding and assembly of the SAFs (at 100 mMin each peptide, 20 1C, pH 7):95 essentially, the two peptides

combine instantly to form the sticky ended dimer; this is

followed by a lag phase on the order of tens of minutes during

Fig. 6 Self-assembly and functionalisation of peptide nanotubes. (a) Diphenylalanine assembles into peptide nanotubes. (b) Schematic of metal

nanowires and coaxial cables templated from these nanotubes. (c) SEM image of vertically aligned peptide nanotubes useful for photoptic

applications. Reproduced in part from ref. 73, copyright r the American Chemical Society. (d) TEM image platinum-coated NSGAITIG

nanotubes (adapted from ref. 74). (e) Expression and self-assembly of filamentous protein Z (FtsZ) into nanofibres. (f) TEM image of FtsZ

filaments coated with silver particles. Reproduced in part from ref. 75 with permission of Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim,

copyright r 2010.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3474 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

which aggregates comprisingB8 sticky ended dimers nucleate;

over the next few hours fibrillogenesis occurs both longi-

tudinally and laterally from these nuclei; subsequent fibrillo-

genesis is essentially one-dimensional leading to matured fibres

B50 nm thick and tens of microns in length. In addition, we

have a good working model for the molecular structure and

organisation in the equilibrium, matured fibres.93 This design

concept, together with our understanding of the mechanism of

assembly and the equilibrium structures of fibres, place us in a

strong position to build on and exploit the SAF systems;

specifically, they allow considerable control over assembly,

and present possibilities for introducing additional compo-

nents either at the start of folding and assembly, or during

fibrillogenesis, Fig. 8.

Rather than seeking decoration of the SAFs per se, we

began by investigating the addition of non-standard, or special

peptides, to the standard SAF mixtures in order to alter the

morphology of the fibres. This included introducing T-shaped

peptides and half-peptides connected through flexible linkers

to bring branches and kinks, respectively, to otherwise straight

and stiff fibres.96–98 One surprise from this early research

was that the special peptides incorporated into the fibres only

Fig. 7 Using M13 bacteriophage virus to create ordered nanostructures. (a) Nonocrystalline nickel, rhodium and cerium templated on M13 for

production of hydrogen (adapted from ref. 82). (b) Schematic and TEM images of M13-templated gold and silver nanowires. Reproduced in part

from ref. 83 with permission of the American Chemical Society, copyrightr 2010. (c) SEM image of Kevlar fibres coated with M13 phage bearing

a peptide that interacts with gold nanoparticles (reproduced in part from ref. 84 with permission of WILEY-VCH Verlag GmbH & Co. KGaA,

Weinheim, copyright r 2007.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3475

sparingly; in some cases although stoichiometric mixtures of the

two parent peptides and the specials used to initiate fibrillo-

genesis (that is, 1 : 1 : 1 mixtures), the resulting fibres incorporate

special peptides to massively sub-stoichiometric levels. We

understand this now from our structural studies:93 the indivi-

dual SAFs are crystalline, and as such are not permissive of

incorporating specials that are significantly different from the

parent peptides; in other words, the introduction of special

peptides is like introducing defects into a crystal lattice. None-

theless, these new peptides do insert into, and do cause the

desired, albeit limited, morphological changes.

Furthermore, using this strategy, we were also able to

incorporate peptides harbouring biotin and FLAG-peptide

tags into the fabric of the SAFs.99 These allowed the sub-

sequent recruitment of streptavidin and anti-FLAG antibodies,

respectively, which could be visualised using nanogold probes

and TEM. Finally in this approach, we have demonstrated

that fluorophore-labelled peptides assembled into the fibres.100

This facilitates direct visualisation of the fibres by light

microscopy (LM), which, in turn, has led us to explore the

epitaxial (polar) growth of the fibres,100 and the late stages of

the kinetics of assembly.95 In these cases, fluorophores can be

introduced after peptide synthesis via bioconjugation, or

during peptide synthesis using modified amino acids. One note

of caution, however, because the SAFs carry an overall

positive charge they recruit negatively charged fluorescein

non-specifically; therefore, exclusively we use positively

charged rhodamine to label peptides.

As with the aforementioned systems, the SAFs can be

used as templates for the deposition of inorganic materials.

Fig. 8 Functionalisation of Self-Assembling peptide Fibres (SAFs). (a) SAF-p1 and SAF-p2a, which when mixed together in equimolar

concentrations at pH 7.4 assemble to form long straight fibres tens of microns long and tens of nanometres thick. (b) Non-covalent decoration

using short 11-residue peptides (SAF-tags),19 which bind via electrostatic interactions. Here decoration is visualised using rhodamine attached to

the N-terminus of the tag. (c) A click-chemistry approach for post-assembly decoration of fibres incorporating azide side chains. Here decoration is

followed by reacting the surface azide moieties with biotin-alkyne, followed by visualisation with 5 nm streptavidin nanogold.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3476 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

We have demonstrated this by depositing silica on both

standard linear SAFs and some of the morphological

variants.101 Here, and in contrast to amyloid-like and

PA-based systems in particular—where the templating

material has to be removed by calcination102,103—the SAF

templates can be removed under mild, ambient conditions by

proteolysis to leave silica replicas.

Returning to decoration with specific chemical and biological

moieties: most recently we have explored two other routes to

decorating the SAFs (ref. 19 and Mahmoud et al. &Woolfson,

unpublished work). This has been done for several reason:

(1) to tackle the aforementioned problem of low levels of

inclusion of special peptides into SAFs, in particular, to

increase these to stoichiometric incorporation of labels and

the subsequent high degrees of decoration; (2) to improve the

biocompatibility of the labelling; and (3) to begin addressing

issues in temporal and spatial control of functionalisation.

Both methods are post-assembly routes (Route 2 in the above

rationale), and therefore demand mild conditions.

The first approach is entirely non-covalent (Route 2b), with

the aim of engineering small peptides that bind the outer

surface of the matured SAFs.19 The rationale is that, as the

fibres exhibit a high degree of structural order and carry an

overall positive charge, one might be able to select peptides

that bind the surfaces. Our initial attempts used peptide

libraries displayed on M13 phage. This failed for a variety of

reasons related to the generally acknowledged stickiness of

M13 phage, which led to high background binding. However,

this, along with the aforementioned observations with

fluorophores, suggested an approach employing short,

negatively charged, synthetic peptides. Briefly, we find that

peptides of the type DEDEDE bind strongly to the fibres;

neutral peptides, AQAQAQ, bind less weakly; and positively

charged peptides, KRKRKR, do not bind at all. Moreover,

the DE- and AQ-based peptides can be tagged, and used to

recruit cargo to the SAFs, as demonstrated with nanogold

particles visualised by TEM, and with fluorophores visualised

directly in solution by LM, Fig. 8b.19 The advantage of these

so-called SAF-tags is that they could be readily added to target

proteins for recruitment to the SAFs. A disadvantage is the

potential for highly dynamic and non-specific binding with

such simple sequences; though in the studies we have

conducted we see long-lived binding and very little evidence

for non-specific interactions.

Our second approach involves small chemically reactive

groups suitable for biocompatible ‘‘click’’ reactions being

directly introduced into the SAF peptides during synthesis

(Route 2a). The modified peptides are then co-assembled with

standard SAF peptides to produce matured fibres. These can

then be decorated with a variety of organic and peptidic

reagents harbouring a complementary click group. To demon-

strate this, we have synthesised SAF peptides incorporating

azide and alkene side chains. Importantly, and unlike our

earlier approaches,99 the modified peptides incorporate

stoichiometrically with the standard SAF peptides. The

resulting modified SAFs can be functionalised through

conjugation with copper-catalysed azide-alkyne and thiol-ene

click reactions, respectively. This allows high levels of

incorporation of the modified peptides (up to 100%), subsequent

high coverage of the SAFs with cargo to be achieved, and the

proportion of the label to be varied. Moreover, as the afore-

mentioned click reactions are orthogonal and both modified

peptides can be co-incorporated into a single fibre preparation,

dual functionalisation of the SAFs is possible. We have

demonstrated all of these possibilities by incorporating either

or both nanogold and fluorophore labels, Fig. 8c. Of course,

this approach requires the synthesis of non-proteinogenic

peptides, which are not amenable to standard recombinant

DNA expression. However, as the incorporation rates are very

high, relatively small amounts of the modified peptides are

required to get respectable levels of decoration, and, so, these

could be used sub-stoichiometrically with standard peptides;

also, side chains that incorporate groups for the click

reactions, and diazo transfer reagents that work in water are

becoming more available.104 Therefore, in our view, this is the

most promising approach to making functionalised SAFs that

we have investigated so far.

4b. Decoration of other coiled-coil-based fibrous systems

A number of other groups have presented designs for fibrous

materials based on a-helical, and in particular coiled-coil

building blocks,6,105,8–10 and some of these have decorated

or functionalised their systems. Notably, Kajava and collea-

gues offer a system similar in some respects to the SAFs;

however, it is based on pentameric coiled coils, and the

individual helices ‘‘slip’’ to give sticky ended assemblies, which

then associate end-to-end and side-by-side to form thickened

fibres depending on the conditions. In a modified design they

incorporate the integrin-binding RGD sequence, to render

fibres that support cell growth in culture.106 In another

SAF-related design approach, Dublin and Conticello describe

a sticky ended trimeric coiled coil with buried histidine

residues. In this case, these imidazole side chains coordinate,

and facilitate the preparation of silver nanowires shrouded in

peptide fibre.107

Conclusion

Through this review, we hope to have outlined the current

activities and trends in decorating, or functionalising, soft

biomaterials, in particular fibrous assemblies. We confess that,

whilst we appreciated that there was a strong and growing

base of literature in the area, we hadn’t expected to find the

depth and breadth of new, exciting and varied research that we

did. Indeed, this review, and the literature cited in it, could

have been expanded considerably had we had the time and

space to do so. As a result of the growth of this field and its

associated literature, rather than simply list and review the

papers that we came across—say, chronologically, by materials

type, or research group—we have tried as far as possible to see

past individual papers and onto generalisations that are

emerging. Specifically, we have categorised the work in terms

of routes to functionalising fibrous biomaterials. Clearly there

are limits to this: some examples are not so easily pigeon-

holed, others span more than one route, and, of course, there

will be future ground-breaking pieces of work that challenge

dogma, and force us to reconsider the categorisation.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3477

Nonetheless, we find this analysis useful, and we hope that

others will find it helpful too.

During the course of our reading and writing, it became

clear that common themes, approaches and methods that

represent best practice are emerging in the field. For example:

(1) the exploitation of natural self-assembling systems, or

peptides derived from these often present a good starting

point. However, this can present restrictions in terms of

control over self-assembly, and also asks the question, just

how does one functionalise materials pre- or post-assembly

without interfering with the self-assembly process itself? This

raises the next two key points from our perspective. (2) That

the design or engineering of the self-assembling and functional

components of the system should be orthogonal. And, related

to this, (3) that this orthogonality should be designed or

engineered into the system, or at least considered, from the

outset of materials development. Perhaps the best examples

that illustrate these two points to date are the peptide amphi-

philes from Stupp and colleagues. (4) Rational peptide design

also represents a good approach in these respects, but it does

rely on having good rules that relate sequence, structure and

assembly. Our understanding for certain protein-folding

motifs, such as the coiled coil, is headed in the right direction,

but this is by no means the case universally for protein folding.

Finally (5) there are now many examples of designed and

engineered self-assembled systems, and it is likely that some

will be better suited to appending certain functions and in

specific applications than others. It would seem prudent, to

choose the right tool for the job in this respect, and not to be

wedded to any one material type. Ideally, as the field develops

we would have an open-source or synthetic-biology ethos, and

a toolkit of basic materials with which to build will emerge.

Somewhat related to this, it is quite understandable that

different groups have adopted similar strategies to demon-

strate that they have achieved decoration. The common

methods are ‘‘functionalisation’’ with a fluorophore of some

description followed by visualisation using light microscopy;

or the addition of gold nanoparticles (GNPs) followed by

visualisation using electron microscopy. Clearly, these are the

best options for demonstrating that decoration has been

achieved, but usually they do not represent functionalisations

in themselves; though examples such as the formation of

metallic nanowires following the recruitment of gold nano-

particles are cases where functional materials are being

developed in this way. Nonetheless, true functionalisation

would be to impart some biological activity such as cell-

binding properties onto the scaffolds. Of course, there is strong

research in this endeavour. However, this is and must remain

one of the tenets of both materials science and synthetic

biology: that is, whilst as a community we must develop the

very best understanding of natural and design self-assembling

systems, as materials scientist and/or synthetic biologists, we

must consider applications that put our materials to use.

In the case of soft fibrous biomaterials, the key applications

areas that are being explored here are scaffolds for 3D cell

culture and tissue engineering, and as sacrificial templates for

the organisation of functional, often hard, materials that are

less readily self-assembled. In many respects, the studies that

we have highlighted have laid the groundwork for this

development, and it is now time to translate further this basic

research and deliver functionalised materials for specific real-

life applications. There are many challenges ahead, including

issues of large-scale materials production, biocompatibility of

methods used in functionalisation, and, for tissue engineering,

immunogenicity of the final materials. However, we feel that

these and other issues are being tackled and will be surmounted.

In summary, the current state of basic research in functional

fibrous biomaterials is buoyant, and provides a strong basis

for translation into real-life applications.

References

1 O. S. Makin and L. C. Serpell, FEBS J., 2005, 272, 5950–5961.2 J. L. Jimenez, E. J. Nettleton, M. Bouchard, C. V. Robinson,

C. M. Dobson and H. R. Saibil, Proc. Natl. Acad. Sci. U. S. A.,2002, 99, 9196–9201.

3 J. D. Hartgerink, E. Beniash and S. I. Stupp, Science, 2001, 294,1684–1688.

4 D. N. Woolfson, Adv. Protein Chem., 2005, 70, 79–112.5 E. H. C. Bromley, K. Channon, E. Moutevelis and

D. N. Woolfson, ACS Chem. Biol., 2008, 3, 38–50.6 S. Kojima, Y. Kuriki, T. Yoshida, K. Yazaki and K. Miura, Proc.

Jpn. Acad., Ser. B, Phys. Biol. Sci., 1997, 73, 7–11.7 M. J. Pandya, G. M. Spooner, M. Sunde, J. R. Thorpe,

A. Rodger and D. N. Woolfson, Biochemistry, 2000, 39,8728–8734.

8 S. A. Potekhin, T. N. Melnik, V. Popov, N. F. Lanina,A. A. Vazina, P. Rigler, A. S. Verdini, G. Corradin andA. V. Kajava, Chem. Biol., 2001, 8, 1025–1032.

9 Y. Zimenkov, V. P. Conticello, L. Guo and P. Thiyagarajan,Tetrahedron, 2004, 60, 7237–7246.

10 K. L. Lazar, H. Miller-Auer, G. S. Getz, J. Orgel andS. C. Meredith, Biochemistry, 2005, 44, 12681–12689.

11 H. Dong, S. E. Paramonov and J. D. Hartgerink, J. Am. Chem.Soc., 2008, 130, 13691–13695.

12 D. N. Woolfson and M. G. Ryadnov, Curr. Opin. Chem. Biol.,2006, 10, 559–567.

13 D. N. Woolfson, Biopolymers, 2010, 94, 118–127.14 L. C. Palmer and S. I. Stupp, Acc. Chem. Res., 2008, 41,

1674–1684.15 H. G. Cui, M. J. Webber and S. I. Stupp, Biopolymers, 2010, 94,

1–18.16 Z. H. Nie and E. Kumacheva, Nat. Mater., 2008, 7, 277–290.17 A. Miyachi, T. Takahashi, S. Matsumura and H. Mihara,

Chem.–Eur. J., 2010, 16, 6644–6650.18 A. J. Baldwin, R. Bader, J. Christodoulou, C. E. MacPhee,

C. M. Dobson and P. D. Barker, J. Am. Chem. Soc., 2006, 128,2162–2163.

19 Z. M. Mahmoud and D. N. Woolfson, Biomaterials, 2010, DOI:10.1016/j.biomaterials.2010.06.041.

20 B. L. Nilsson, M. B. Soellner and R. T. Raines, Annu. Rev.Biophys. Biomol. Struct., 2005, 34, 91–118.

21 K. Channon and C. E. MacPhee, Soft Matter, 2008, 4, 647–652.22 O. S. Makin, P. Sikorski and L. C. Serpell, Curr. Opin. Chem.

Biol., 2006, 10, 417–422.23 K. J. Channon, G. L. Devlin and C. E. MacPhee, J. Am. Chem.

Soc., 2009, 131, 12520–12521.24 E. Gazit, Chem. Soc. Rev., 2007, 36, 1263–1269.25 R. V. Ulijn and A. M. Smith, Chem. Soc. Rev., 2008, 37, 664–675.26 T. P. J. Knowles, T. W. Oppenheim, A. K. Buell, D. Y. Chirgadze

and M. E. Welland, Nat. Nanotechnol., 2010, 5, 204–207.27 S. L. Gras, A. K. Tickler, A. M. Squires, G. L. Devlin,

M. A. Horton, C. M. Dobson and C. E. MacPhee, Biomaterials,2008, 29, 1553–1562.

28 A. Horii, X. M. Wang, F. Gelain and S. G. Zhang, PLoS One,2007, 2, e190, DOI: 10.1371/journal.pone.0000190.

29 Y. Kumada and S. G. Zhang, PLoS One, 2010, 5, e10305, DOI:10.1371/journal.pone.0010305.

30 L. Haines-Butterick, K. Rajagopal, M. Branco, D. Salick,R. Rughani, M. Pilarz, M. S. Lamm, D. J. Pochan and

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

3478 Chem. Soc. Rev., 2010, 39, 3464–3479 This journal is c The Royal Society of Chemistry 2010

J. P. Schneider, Proc. Natl. Acad. Sci. U. S. A., 2007, 104,7791–7796.

31 D. A. Salick, J. K. Kretsinger, D. J. Pochan and J. P. Schneider,J. Am. Chem. Soc., 2007, 129, 14793–14799.

32 A. Firth, A. Aggeli, J. L. Burke, X. B. Yang and J. Kirkham,Nanomedicine, 2006, 1, 189–199.

33 S. Kyle, A. Aggeli, E. Ingham and M. J. McPherson, TrendsBiotechnol., 2009, 27, 423–433.

34 S. Kasai, Y. Ohga, M. Mochizuki, N. Nishi, Y. Kadoya andM. Nomizu, Biopolymers, 2004, 76, 27–33.

35 Y. Ohga, F. Katagiri, K. Takeyama, K. Hozumi, Y. Kikkawa,N. Nishi and M. Nomizu, Biomaterials, 2009, 30, 6731–6738.

36 K. Channon, E. H. C. Bromley and D. N. Woolfson, Curr. Opin.Struct. Biol., 2008, 18, 491–498.

37 U. Baxa, V. Speransky, A. C. Steven and R. B. Wickner, Proc.Natl. Acad. Sci. U. S. A., 2002, 99, 5253–5260.

38 S. C. Kundu, B. C. Dash, R. Dash and D. L. Kaplan, Prog.Polym. Sci., 2008, 33, 998–1012.

39 A. Leal-Egana and T. Scheibel, Biotechnol. Appl. Biochem., 2010,55, 155–167.

40 E. Bini, C. W. P. Foo, J. Huang, V. Karageorgiou, B. Kitchel andD. L. Kaplan, Biomacromolecules, 2006, 7, 3139–3145.

41 A. W. Morgan, K. E. Roskov, S. Lin-Gibson, D. L. Kaplan,M. L. Becker and C. G. Simon, Biomaterials, 2008, 29,2556–2563.

42 C. M. Elvin, A. G. Carr, M. G. Huson, J. M. Maxwell,R. D. Pearson, T. Vuocolo, N. E. Liyou, D. C. C. Wong,D. J. Merritt and N. E. Dixon, Nature, 2005, 437, 999–1002.

43 P. S. Nadaud, M. Sarkar, B. Wu, C. E. MacPhee, T. J. Maglieryand C. P. Jaroniec, Protein Expression Purif., 2010, 70, 101–108.

44 D. A. Harrington, E. Y. Cheng, M. O. Guler, L. K. Lee,J. L. Donovan, R. C. Claussen and S. I. Stupp, J. Biomed. Mater.Res., Part A, 2006, 78a, 157–167.

45 M. A. Greenfield, J. R. Hoffman, M. O. de la Cruz andS. I. Stupp, Langmuir, 2010, 26, 3641–3647.

46 R. N. Shah, N. A. Shah, M. M. D. Lim, C. Hsieh, G. Nuber andS. I. Stupp, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 3293–3298.

47 T. D. Sargeant, S. M. Oppenheimer, D. C. Dunand andS. I. Stupp, J. Tissue Eng. Regener. Med., 2008, 2, 455–462.

48 A. Mata, Y. Geng, K. J. Henrikson, C. Aparicio, S. R. Stock,R. L. Satcher and S. I. Stupp, Biomaterials, 2010, 31, 6004–6012.

49 S. Ghanaati, M. J. Webber, R. E. Unger, C. Orth, J. F. Hulvat,S. E. Kiehna, M. Barbeck, A. Rasic, S. I. Stupp andC. J. Kirkpatrick, Biomaterials, 2009, 30, 6202–6212.

50 L. Aulisa, H. Dong and J. D. Hartgerink, Biomacromolecules,2009, 10, 2694–2698.

51 K. M. Galler, L. Aulisa, K. R. Regan, R. N. D’Souza andJ. D. Hartgerink, J. Am. Chem. Soc., 2010, 132, 3217–3223.

52 I. C. Reynhout, J. Cornelissen and R. J. M. Nolte, J. Am. Chem.Soc., 2007, 129, 2327–2332.

53 T. B. Yu, J. Z. Bai and Z. B. Guan, Angew. Chem., Int. Ed., 2009,48, 1097–1101.

54 A. J. Dirks, S. S. van Berkel, H. I. V. Amatdjais-Groenen,F. Rutjes, J. Cornelissen and R. J. M. Nolte, Soft Matter, 2009,5, 1692–1704.

55 E. K. Schillinger, E. Mena-Osteritz, J. Hentschel, H. G. Bornerand P. Bauerle, Adv. Mater., 2009, 21, 1562–1567.

56 M. A. Gauthier and H. A. Klok, Chem. Commun., 2008,2591–2611.

57 H. G. Borner, Prog. Polym. Sci., 2009, 34, 811–851.58 H. R. Marsden and A. Kros,Macromol. Biosci., 2009, 9, 939–951.59 J. M. Baskin and C. R. Bertozzi, Aldrichimica Acta, 2010, 43,

15–23.60 C. A. DeForest, B. D. Polizzotti and K. S. Anseth, Nat. Mater.,

2009, 8, 659–664.61 H. Kodama, S. Matsumura, T. Yamashita and H. Mihara, Chem.

Commun., 2004, 2876–2877.62 S. Matsumura, S. Uemura and H. Mihara, Chem.–Eur. J., 2004,

10, 2789–2794.63 A. Y. Wang, X. Mo, C. S. Chen and S. M. Yu, J. Am. Chem. Soc.,

2005, 127, 4130–4131.64 X. Mo, Y. J. An, C. S. Yun and S. M. Yu, Angew. Chem., Int. Ed.,

2006, 45, 2267–2270.65 H. J. Lee, J. S. Lee, T. Chansakul, C. Yu, J. H. Elisseeff and

S. M. Yu, Biomaterials, 2006, 27, 5268–5276.

66 A. Y. Wang, S. Leong, Y. C. Liang, R. C. C. Huang, C. S. Chenand S. M. Yu, Biomacromolecules, 2008, 9, 2929–2936.

67 L. S. Li and S. I. Stupp, Angew. Chem., Int. Ed., 2005, 44,1833–1836.

68 M. O. Guler, S. Soukasene, J. F. Hulvat and S. I. Stupp, NanoLett., 2005, 5, 249–252.

69 M. B. Dickerson, K. H. Sandhage and R. R. Naik, Chem. Rev.,2008, 108, 4935–4978.

70 L. L. del Mercato, P. Rivera-Gil, A. Z. Abbasi, M. Ochs,C. Ganas, I. Zins, C. Sonnichsen and W. J. Parak, Nanoscale,2010, 2, 458–467.

71 D. J. Evans, Inorg. Chim. Acta, 2010, 363, 1070–1076.72 M. Reches and E. Gazit, Science, 2003, 300, 625–627.73 N. Amdursky, M. Molotskii, D. Aronov, L. Adler-Abramovich,

E. Gazit and G. Rosenman, Nano Lett., 2009, 9, 3111–3115.74 E. Kasotakis, E. Mossou, L. Adler-Abramovich, E. P. Mitchell,

V. T. Forsyth, E. Gazit and A. Mitraki, Biopolymers, 2009, 92,164–172.

75 N. Ostrov and E. Gazit, Angew. Chem., Int. Ed., 2010, 49,3018–3021.

76 T. Scheibel, R. Parthasarathy, G. Sawicki, X. M. Lin, H. Jaegerand S. L. Lindquist, Proc. Natl. Acad. Sci. U. S. A., 2003, 100,4527–4532.

77 S. W. Lee, C. B. Mao, C. E. Flynn and A. M. Belcher, Science,2002, 296, 892–895.

78 K. T. Nam, Y. J. Lee, E. M. Krauland, S. T. Kottmann andA. M. Belcher, ACS Nano, 2008, 2, 1480–1486.

79 Y. Huang, C. Y. Chiang, S. K. Lee, Y. Gao, E. L. Hu, J. DeYoreo and A. M. Belcher, Nano Lett., 2005, 5, 1429–1434.

80 Y. J. Lee, H. Yi, W. J. Kim, K. Kang, D. S. Yun, M. S. Strano,G. Ceder and A. M. Belcher, Science, 2009, 324, 1051–1055.

81 Y. S. Nam, T. Shin, H. Park, A. P. Magyar, K. Choi, G. Fantner,K. A. Nelson and A. M. Belcher, J. Am. Chem. Soc., 2010, 132,1462–1463.

82 B. Neltner, B. Peddie, A. Xu, W. Doenlen, K. Durand, D. S. Yun,S. Speakman, A. Peterson and A. Belcher, ACS Nano, 2010, 4,3227–3235.

83 Y. J. Lee, Y. Lee, D. Oh, T. Chen, G. Ceder and A. M. Belcher,Nano Lett., 2010, 10, 2433–2440.

84 C. Y. Chiang, C. M. Mello, J. J. Gu, E. Silva, K. J. Van Vliet andA. M. Belcher, Adv. Mater., 2007, 19, 826–832.

85 O. Carny, D. E. Shalev and E. Gazit, Nano Lett., 2006, 6,1594–1597.

86 N. Ostrov and E. Gazit, Angew. Chem., Int. Ed., 2010, 49,3018–3021.

87 D. Gottlieb, S. A. Morin, S. Jin and R. T. Raines, J. Mater.Chem., 2008, 18, 3865–3870.

88 L. C. Palmer, C. J. Newcomb, S. R. Kaltz, E. D. Spoerke andS. I. Stupp, Chem. Rev., 2008, 108, 4754–4783.

89 E. D. Spoerke, S. G. Anthony and S. I. Stupp, Adv. Mater., 2009,21, 425–430.

90 H. W. Jun, V. Yuwono, S. E. Paramonov and J. D. Hartgerink,Adv. Mater., 2005, 17, 2612–2617.

91 J. Kirkham, A. Firth, D. Vernals, N. Boden, C. Robinson,R. C. Shore, S. J. Brookes and A. Aggeli, J. Dent. Res., 2007,86, 426–430.

92 A. M. Smith, E. F. Banwell, W. R. Edwards, M. J. Pandya andD. N. Woolfson, Adv. Funct. Mater., 2006, 16, 1022–1030.

93 D. Papapostolou, A. M. Smith, E. D. T. Atkins, S. J. Oliver,M. G. Ryadnov, L. C. Serpell and D. N. Woolfson, Proc. Natl.Acad. Sci. U. S. A., 2007, 104, 10853–10858.

94 E. F. Banwell, E. S. Abelardo, D. J. Adams, M. A. Birchall,A. Corrigan, A. M. Donald, M. Kirkland, L. C. Serpell,M. F. Butler and D. N. Woolfson, Nat. Mater., 2009, 8, 596–600.

95 E. H. C. Bromley, K. J. Channon, P. J. S. King, Z. N. Mahmoud,E. F. Banwell, M. F. Butler, M. P. Crump, T. R. Dafforn,M. R. Hicks, J. D. Hirst, A. Rodger and D. N. Woolfson,Biophys. J., 2010, 98, 1668–1676.

96 M. G. Ryadnov and D. N. Woolfson, Angew. Chem., Int. Ed.,2003, 42, 3021–3023.

97 M. G. Ryadnov and D. N.Woolfson,Nat.Mater., 2003, 2, 329–332.98 M. G. Ryadnov and D. N. Woolfson, J. Am. Chem. Soc., 2005,

127, 12407–12415.99 M. G. Ryadnov and D. N. Woolfson, J. Am. Chem. Soc., 2004,

126, 7454–7455.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online

This journal is c The Royal Society of Chemistry 2010 Chem. Soc. Rev., 2010, 39, 3464–3479 3479

100 A. M. Smith, S. F. A. Acquah, N. Bone, H. W. Kroto,M. G. Ryadnov, M. S. P. Stevens, D. R. M. Walton andD. N. Woolfson, Angew. Chem., Int. Ed., 2005, 44, 325–328.

101 S. C. Holmstrom, P. J. S. King, M. G. Ryadnov, M. F.Butler, S. Mann and D. N. Woolfson, Langmuir, 2008, 24,11778–11783.

102 J. E. Meegan, A. Aggeli, N. Boden, R. Brydson, A. P. Brown,L. Carrick, A. R. Brough, A. Hussain and R. J. Ansell, Adv.Funct. Mater., 2004, 14, 31–37.

103 V. M. Yuwono and J. D. Hartgerink, Langmuir, 2007, 23,5033–5038.

104 S. F. M. van Dongen, R. L. M. Teeuwen, M. Nallani, S. S. vanBerkel, J. Cornelissen, R. J. M. Nolte and J. C. M. van Hest,Bioconjugate Chem., 2009, 20, 20–23.

105 D. E. Wagner, C. L. Phillips, W. M. Ali, G. E. Nybakken,E. D. Crawford, A. D. Schwab, W. F. Smith and R. Fairman,Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 12656–12661.

106 V. Villard, O. Kalyuzhniy, O. Riccio, S. Potekhin, T. N. Melnik,A. V. Kajava, C. Ruegg and G. Corradin, J. Pept. Sci., 2006, 12,206–212.

107 S. N. Dublin and V. P. Conticello, J. Am. Chem. Soc., 2008, 130,49–51.

Publ

ishe

d on

02

Aug

ust 2

010.

Dow

nloa

ded

by U

nive

rsity

of

Bri

stol

on

12/1

0/20

15 1

4:30

:42.

View Article Online