Knight_et_al-2015-Advanced_Materials

REVIEW

S

equence Programmable Peptoid Polymers for Diverse Materials Applications

A bigail S. K night ,E f?e Y. Z hou ,M atthew B. F rancis ,a nd R onald N . Z uckermann*

A . S. Knight, E. Y . Zhou, Prof. M. B. Francis

U C Berkeley Chemistry Department L atimer Hall ,B erkeley ,C A 94720 ,U SA

D r. R. N. Zuckermann, Prof. M. B. Francis T he Molecular Foundry Lawrence Berkeley National Lab 1Cyclotron Road ,B erkeley ,C A 94720 ,U SA E-mail: r nzuckermann@https://www.wendangku.net/doc/a04767916.html, DOI: 10.1002/adma.201500275

1. I ntroduction

T he challenge of creating synthetic materials with the structural sophistication and complex function found in biology has been a long-term goal in materials science ( F igure 1). Research from

both the biological and chemical communities has been con-verging to ? ll the large gap between synthetic homopolymers and native biological materials. Nature has evolved a variety of sequence programmable polymers that have functions ranging from carrying the genetic code to catalyzing chemical reac-tions. Natural biopolymers represent the pinnacle of sequence control, and this has allowed for the evolution of a plethora of sophisticated functions. These functions are enabled by the P olymer sequence programmability is required for the diverse structures and complex properties that are achieved by native biological polymers, but

efforts towards controlling the sequence of synthetic polymers are, by com-parison, still in their infancy. Traditional polymers provide robust and chemi-cally diverse materials, but synthetic control over their monomer sequences is limited. The modular and step-wise synthesis of peptoid polymers, on the other hand, allows for precise control over the monomer sequences, affording

opportunities for these chains to fold into well-de? ned nanostructures. Hun-dreds of different side chains have been incorporated into peptoid polymers

using ef? cient reaction chemistry, allowing for a seemingly in? nite variety of possible synthetically accessible polymer sequences. Combinatorial discovery techniques have allowed the identi? cation of functional polymers within large libraries of peptoids, and newly developed theoretical modeling tools

speci? cally adapted for peptoids enable the future design of polymers with desired functions. Work towards controlling the three-dimensional structure

of peptoids, from the conformation of the amide bond to the formation of protein-like tertiary structure, has and will continue to enable the construc-tion of tunable and innovative nanomaterials that bridge the gap between

natural and synthetic polymers.

precision control of molecular interactions on the ?ngstr?m scale – the length scale

of individual bonds. Proteins have com-plex networks of noncovalent interactions that are necessary for speci? c molecular

recognition (e.g. binding to speci? c DNA

sequences), as well as precisely placed functional groups that are required for

catalysis. These exquisite nanoscale archi-tectures are typically folded from abso-lutely monodisperse linear polymer chains of a precisely de? ned monomer sequence.

Biological research has been moving towards the production of novel materials using the natural biosynthetic machinery;

recent work in directed protein evolution [ 1] and the incorporation of unnatural amino acids

[ 2–4 ] has made it possible to evolve proteins for non-native functions. How-ever, biopolymers and materials based on proteins or DNA suffer from inherent lim-itations, including the necessity of water as their solvent, susceptibility to enzymatic degradation, and poor environmental sta-bility to variation in temperature, pH, and ionic strength.

T he study of synthetic polymeric materials began with homo-polymers and copolymers, and has expanded into functional

heteropolymers. While polymeric materials tend to be signi? -cantly more robust than protein-based materials, they lack the

structural sophistication achieved by proteins. In recent years, polymer chemists have sought to improve synthesis methods to emulate biology’s degree of sequence control, and the gap

between synthetic polymers and native biological materials is

beginning to close.

[ 5–7 ] Many synthetic strategies for polymer sequence control have recently been developed, including tem-plating, kinetic control, orthogonal reactivities, and the linking

of structures after polymerization.

[ 5] These techniques have allowed for signi? cant advances in synthetic polymer materials,

especially in the development of novel photonics and dynamic

materials.

[ 7–9 ] The demonstrated utility of partial sequence con-trol illuminates the abundance of opportunities available as the synthetic tools for precise monomer control continue to improve. S tep-wise synthesis provides the opportunity to generate polymers with complete sequence control. This strategy was pioneered using solid supports by Merri? eld for peptide syn-thesis, [ 10 ] but has also been developed for the synthesis of

DNA, RNA and a variety of resin-bound organic structures. [ 11 ]

R E V I E W

However, the extension of these techniques to create sequence-de? ned non-natural polymers of signi? cant main chain lengths has been a longstanding challenge. The monomer coupling reactions used to make such polymers need to have nearly quantitative yields in order to synthesize large, well-de? ned polymer chains. Peptidomimetic N -substituted glycine poly-mers, or peptoids, break this yield barrier and have thus opened up a new class of sequence-de? ned polymers ( F igure 2). [ 12 ]Pep-toid polymers thus straddle the boundary between biological materials and synthetic polymers: their biomimetic structure and precision sequence control provides a breadth of oppor-tunities to produce sequence programmable, folded polymers with the robustness typical of synthetic materials. T he step-wise solid-phase synthesis of peptoids was devel-oped in 1992, [ 13 ] and since then their application to various ? elds has grown exponentially. This synthesis strategy provides access to an almost in? nite diversity of polymer sequences, since each monomer is individually tunable. The large popu-lation of potential polymers opens the opportunity to screen for particular sequences with new functionalities using com-binatorial chemistry.

[ 14 ] Moreover, the ability to identify func-tional structures from such libraries accelerates the discovery of materials with advanced properties. Although this capability exists with genetically-speci? ed biological polymers (e.g., poly-peptides and polynucleotides), peptoids are a more versatile option due to their increased chemical diversity, increased sta-

bility to degradative enzymes and environmental variables, and

A bigail Knight received her B.S. in chemistry from the University of North Carolina, Chapel Hill in 2010 and is currently ? nishing her Ph.D at the University of California, Berkeley under the guidance of Professor Matthew

Francis. After the completion of her Ph.D. she will begin a postdoctoral position at the University of California, Santa Barbara with Professor Craig Hawker. Her research interests include bioinspired materials and selective metal coordination. E f? e Zhou recently completed her B.S. in chemistry at the University of California at Berkeley working in the labora-tory of Matthew Francis. She is interested in working at the interface of chemistry and biology in her future research career.

M atthew Francis received his B.S. in Chemistry from Miami University, and he received his Ph.D. from Harvard University working with Prof. Eric Jacobsen. He was a Postdoctoral Fellow with Prof. Jean Fréchet through the Miller Institute for Basic Research in Science at UC Berkeley, and he started his independent career in the UC Berkeley Chemistry Department in 2001. He has built a research program involving the development of new reac-tions for protein modi? cation. These new chemical tools have then been used to generate complex biomolecular assemblies for use in in vivo delivery, light harvesting, and water treatment applications.

R onald Zuckermann is a Sr. Scientist and Facility Director of the Biological Nanostructures Facility at the Molecular Foundry at the Lawrence Berkeley National Laboratory. He obtained a Ph.D. in Chemistry from UC Berkeley in 1989, and worked as a research scientist in the biotech-nology industry for 16 years developing combinatorial drug discovery technologies. He was named a Chiron Research Fellow in 2003, and an LBNL Sr. Scientist in 2011. He invented sequence-de? ned peptoid polymers, and works on folding them into precise nanoscale architectures. He adapts structural design rules from biology and applies them to the world of materials science.

F rom left to right: R.Z., M.F., A.K., and E.Z.

F igure 2. P eptide versus peptoid structure. Example tetramers are shown

of each.

F igure 1. T he range of characteristics for synthetic and biological poly-mers reveals a gap that is strategically ? lled by peptoid-based materials.

REVIEW

a decreased synthetic cost. The step-wise synthesis of peptoid polymers makes the nearly absolute monodispersity encoun-tered with biological polymers accessible, allowing for unique structural control in a synthetic polymer. Recent reviews have

described applications of peptoids as small molecules [ 15–17 ]

and polymers, [ 18–22 ] and in the ? eld of biomaterials. [ 12,23 ]This

review aims to highlight the unique applications of peptoids as modular and tunable structures with the potential to bridge the gap between synthetic and biological polymers in the future of bioinspired materials research.

2. S ynthesis and Structural Control of Peptoids

2.1. S ynthesis

P eptoid polymers have been synthesized using two strategies: an iterative step-wise technique, typically performed on a solid support, that allows precise control of sequence de? nition, and a polymerization approach, typically completed in solution, that allows access to higher molecular weight polymers. The step-wise submonomer synthesis uses a two-step monomer addition cycle, similar to solid-phase peptide synthesis ( F igure 3a ). [ 24 ] Both reactions used in the submomoner synthesis strategy can achieve near quantitative yields, allowing the ef? cient synthesis of virtually monodisperse sequence-de? ned polymers. This syn-thesis, which can be performed manually or by an automated synthesizer, can yield polymers of up to ? fty monomer units in length. For larger polymers, the ring-opening polymerization (ROP) of N -s ubstituted α-amino acid- N -carboxyanhydrides ( N -substituted NCAs) has been applied to create peptoid homo-polymers and copolymers (Figure 3 b ).

[ 22 ] Less sequence con-trol is available with the polymerization approach, but a large number of different side chains have been incorporated, pro-viding access to a wide variety of polymer structures within the same backbone. T o access sequence-de? ned peptoid polymers, a step-wise modular synthesis is required. Although the iterative synthesis appears tedious and time consuming in comparison, both syn-thesis on a solid-support and the automation of the coupling reactions enables rapid and facile synthesis of these polymers. The nearly quantitative yields of the two-step synthesis lead to high yields of very pure polymer material. The modular syn-thesis begins with the acylation of a resin-bound amine with a haloacetic acid, followed by displacement of the halogen with a primary amine (Figure 3 a ). The nature of this synthesis allows for the use of diverse functional groups as side chains, since nearly any primary amine can be incorporated as a submon-omer. Unlike the synthetic strategies used for the synthesis of biological polymers, this synthetic scheme does not require the synthesis of special building blocks. Correspondingly, hun-dreds of side chains can be easily incorporated at each position directly from commercially available materials. Side chains with reactive functional groups necessitate protection, but many het-erocycles can be used directly with no protection. [ 25 ]

S ince the ? rst demonstration of the submonomer synthesis

technique, [ 24 ] hundreds of monomers including ionic, aro-matic, heterocyclic, and aliphatic moieties have been incorpo-rated. [ 26 ] Almost any primary amine can perform the halogen

displacement, providing an unlimited library of functional groups. Functional groups that have been incorporated on the side chains range from small aliphatic groups, to amino acid-like side chains, to larger moieties (e.g., carbohydrates, dyes, and cofactors) and individual side chains have been demon-strated to contribute signi? cantly to polymer properties and function. A few examples include azobenzene side chains that

lead to photoresponsive polymers [ 27 ] and chiral monomers that

introduce entropic constraints and promote the formation of

F igure 3. A n outline of the most common techniques used for the syn-thesis of α-peptoids. a) Peptoid synthesis techniques used with solid-phase polystyrene resins. Carbodiimides are most commonly used as the

coupling agent and piperidine is commonly used for the deprotection step. b) Solution polymerization using N -substituted NCAs and NTAs for peptoid polymer synthesis. For RE(BH 4)3(THF) 3 , the rare earth (RE) can be Sc, Y , La, Nd, Dy, or Lu.

R E V I E W

secondary structures. [ 28 ]

The submonomer synthesis method has also been expanded to include chemoselective coupling partners, such as ketones, aldehydes, and thiols, for conjuga-tion of peptoid polymers to other molecules or to substrates. [ 29 ]

Additionally, Kirshenbaum et al. demonstrated the incorpora-tion of larger moieties, including hormones and metal com-plexes, which were added to a fully elongated peptoid using

click chemistry. [ 30 ] Their work compared the electrochemical

properties of ferrocene-peptoid conjugates to the properties of free ferrocene, noting a minimal attenuation of electrochemical properties. W hile traditional synthetic techniques allow for rapid and facile synthesis of peptoid polymers, methods relying on bio-synthetic pathways could increase the ease of synthesis and enable larger sequence-de? ned peptoid polymers to be formed. The current techniques only produce small oligomers, but these strategies have huge potential. A ribosomal synthesis has been described for the synthesis of peptoids and peptide-peptoid hybrids by the Suga group. [ 31 ] A cell-free translation

system was used [ 32 ] in addition to ? exizymes – arti? cial ami-noacyl-tRNA ribozymes [ 33 ] – allowing for the incorporation of

N -substituted glycines. A variety of functional groups were evaluated, including branched and unbranched alkyl chains and functional chains such as esters, azides, alkenes, and alkynes. Although the yields were low, linear peptoids of up to six monomers were synthesized in addition to cyclic peptide-peptoid hybrids linked with a thioether. Biological machinery has also been applied to the synthesis β-peptoid homopoly-mers; a lipase was employed to catalyze the synthesis of poly( N -(2-hydroxyethyl)-β-propylamide. [ 34 ] The synthesis of hexamers was characterized by MALDI-TOF MS, and the structure was further derivatized to form a brush with polycaprolactone. The copolymer was cast to form ? lms of microstructures character-ized by SEM. Although all of the peptoids that have been syn-thesized with biological machinery are limited in length, these groups have opened up the avenue for enzyme-based synthesis of peptoid-polymers. A variety of combinatorial platforms using the submonomer synthesis have been developed and employed for identifying functional peptoid sequences. Libraries were initially developed for drug discovery applications such as ? nding low molecular

weight structures that bind speci? cally to a protein. [ 35 ]They

have since been used for additional applications including

identifying antimicrobial

[ 36 ] and antifouling [ 37 ] molecules and ligands for selective metal coordination.

[ 38 ] The strength of the combinatorial approach is the ability to screen thousands of molecules for a particular functionality, but once the active compound is identi? ed the structure must be removed from the resin to identify the sequence. Therefore, a cleavable linker such as a methionine (cleavable with CNBr) or a photocleavable linker must be incorporated. Once the peptoid is cleaved, the structure is typically identi? ed using tandem mass spectrom-etry (MS-MS). In particular, the Kodadek group has contributed to the body of methods for synthesis and screening of combi-natorial peptoid libraries. Techniques that could be applied to

materials applications include the design of dimeric [ 39 ]and

cyclic [ 40 ] peptoid libraries that can be sequenced with MS-MS and the development of a two-color quantum dot-based

screen. [ 39 ] Their work has additionally highlighted the utility of

redundant combinatorial libraries to eliminate false positives

before post-screening validation. [ 41 ] Combinatorial techniques

have been applied not only on polystyrene-based resins, but

additionally on glass slides [ 42 ] and cellulose; [ 43 ] this diversity of

substrates allows for the development of novel materials such as microarrays that have been used to identify substrates for the

early detection of diseases. [ 44 ]

W hile step-wise synthesis is required for sequence control of peptoid polymers and enables the application of combinatorial chemistry, there are still limitations in the length of the poly-mers that can be accessed. Techniques using traditional living polymerizations have been developed to overcome this limita-tion. The ring-opening polymerization of N -substituted NCAs

to form polysarcosine was initially performed in the 1926. [ 45 ]In

more recent research efforts, the Luxenhofer group has synthe-sized a variety of polypeptoids using the ROP of N -substituted NCAs (Figure 3 b ). They have demonstrated the reliability of living polymerization by making well controlled homopolymers and multiblock copolymers. These copolymers are not acces-sible via the ROP of NCAs without the N -substitution due to

irreversible chain termination. [ 46 ] Polymerization of N -substi-tuted NCAs leads to high monomer conversion and yields poly-mers with molecular weights of approximately 1 – 10 kg/mol

[ 47 ] and low polydispersity index ( D M ) values (below 1.1).

[ 48 ]The polymers are ? exible and have a diversity of side chains. Ali-phatic side chains with one to four carbons have been evaluated; polymers with shorter side-chains were determined to be water-soluble and those with longer chains were insoluble in water. This allowed for the design of amphiphilic block copolymers. The polymers were characterized by matrix-assisted laser des-orption/ionization time of ? ight mass spectrometry (MALDI-TOF MS), NMR, and GPC. The kinetics of the polymerization were determined to be pseudo ? rst-order with respect to the ini-tiator, allowing for predictable molecular weights. A few variants of the ring-opening polymerization of N -sub-stituted NCAs have also been explored. Zhang and coworkers have described the polymerization of N -substituted NCAs using N -heterocyclic carbene-mediated ROP (Figure 3 b ), which inter-estingly can yield either linear or cyclic polymers. [ 49 ] The syn-thesis of the cyclic polymers was determined to proceed with pseudo-? rst order kinetics and lead to high monomer conver-sions with narrow weight distributions ( D M = 1.03 – 1.15). [ 50 ] Cyclic poly( N -decylglycine)s were synthesized with an impres-sive range of molecular weights (4.8 – 31 kg/mol). Block copoly-mers were synthesized with poly( N -methylglycine) achieving similar monomer conversion yields. These block-copolymers were demonstrated to form spherical micelles that transitioned in tubular micelles over time. Ling et al. used a N -substituted glycine N -thiocarboxyanhydrides ( N -substituted NTA), a thio-variant of the N -substituted NCAs, to synthesize hydrophilic

polymers with sarcosine monomers,

[ 51 ] hydrophobic polymers with butyl side-chains, and block copolymers which are ther-moresponsive. [ 52 ] The N -substituted NTAs are simpler to syn-thesize than the corresponding N -substituted NCAs and are

stable for up to a year at room temperature in an inert atmos-phere. Polymerizations of N -substituted NTAs required higher temperatures due to lower reactivity, however, the yields, molec-ular weights, and polydispersities of the synthesized polymers are comparable to those polymerized from N -substituted NCAs.

REVIEW

C ritical to the development of functional peptoid-based mate-rials is the ability to immobilize peptoid polymers on various surfaces. There are two fundamental approaches to this: syn-thesizing polymers from a substrate surface, and ligating struc-tures to a surface after synthesis. Examples of synthesizing polymers directly from a substrate include synthesis on glass

surfaces [ 42,53 ] and living polymerization on a 1% crosslinked

polystyrene beaded solid-support, which is commonly used

for solid-phase synthesis. [ 47 ] Photolithographic synthesis has

been reported using UV-irradiation to unmask a protected sur-face moiety, allowing spatial and density control over step-wise

peptoid synthesis on a glass surface. [ 42 ] Additionally, amphi-philic block polypeptoids were synthesized using surface initi-ated polymerization from a glass surface. [ 53 ] The height of the

polymer brushes formed was characterized by atomic force microscopy, and the surface was characterized using contact angle measurements. Multiblock copolymers have also been synthesized on a polystyrene resin using ROP of N -s ubstituted

NCAs, a polymerization typically used in solution. [ 47 ] Up to

? ve blocks of polymers with small side chains and low kg/mol molecular weights have been synthesized on a polystyrene resin, then cleaved and characterized with MALDI-TOF MS and 1 H -NMR. A strong Br?nsted acid (HBF 4 ) was used to decrease the polydispersity ( D M = 1.44) by slowing the polymerization. This synthesis is rapid and cost-ef? cient and is particularly promising due to the potential to be used in combination with submonomer peptoid synthesis or solid-phase peptide syn-thesis to create hybrid structures with unique bioconjugation capabilities and various applications.

2.2. M odi? ed Primary Structure

S ide chain diversity enables structural and functional con-trol, but additional conformations and applications could be accessed by backbone-variants of the traditional α-peptoid. The most common variant of the N -substituted glycine peptoid

backbone is the β-peptoid, or N -alkyl-β-alanine oligomer.

[ 54 ]A variety of synthetic routes have been applied to the synthesis of these structures ( F igure 4 a ), including multiple methods of solution phase polymerization. The ? rst catalytic synthesis

of poly-β-peptoids was reported by Hanton and coworkers. [ 55 ]

This copolymerization of N -alkylaziridines and carbon mon-oxide was used to synthesize poly-β-peptoids with quantitative yields and controlled molecular weights (2 – 11 kg/mol with D M = 1.11 – 1.64). The reaction is catalyzed by BnCOCo(CO) 4,which is readily synthesized and puri? ed with an extraction before polymerization. The polymers have been characterized by 1 H and 13 C -NMR in addition to IR and MALDI-TOF MS. This work was continued by Li J ia's group who investigated

the mechanism of the polymerization with various catalysts.

[ 56 ] Using the aforementioned technique, multiple side-products can form, including a lactam and polyamines instead of the amide based-polymer. To gain a more comprehensive under-standing of the mechanism, in situ infrared spectroscopy was used to monitor chain growth over time. One of the catalysts, CH 3C (O)Co(CO) 3P ( o -tolyl) 3 , allowed rapid dissociation of the P( o -tolyl) 3 ligand, and the remaining cobalt complex ef? ciently performed the polymerization without catalyzing the forma-

tion of a lactam side-product. β-peptoids synthesized using cobalt-based catalysts were characterized using NMR and MALDI-TOF MS, and subsequently immobilized on gold as

an anti-fouling material.

[ 57 ] A similar technique involving the cobalt-catalyzed polymerization of N -alkylazetidine and carbon monoxide has been used to synthesize γ-peptoids, which con-tain one additional methylene in the backbone compared to the

β-peptoid.

[ 58,59 ] This synthesis does not have as high yield as those for β-peptoids due to competing reactions, and therefore the applications of these peptoid variants are rare. S equence-de? ned β-peptoids have also been synthesized using step-wise chemistry (Figure 4 a ). The ? rst approach used a two-step monomer addition cycle involving acylation with

F igure 4. A n outline of peptoid backbone variants. a) β-peptoid and γ-peptoid structures and synthesis techniques. b) Additional variations on the α-peptoid backbone structures. c) Branching peptoid-based struc-tures. Various side chains were incorporated into the podand, and ? rst, second, and third generation dendrimers were synthesized using the structures on the right.

R E V I E W

an acrylic acid followed by a Michael addition with a primary

amine. [ 60 ] Further work was done by Arvidsson et al. opti-mizing the synthesis and exploring Lewis acid catalysts. [ 61 ]They obtained the highest yields using Tentagel S PHB as a solid-sup-port and a solvent system of water and tetrahydrofuran (THF). In this work, a nucleophilic submonomer synthesis similar to that used for α-peptoids was also evaluated and determined to be less ef? cient. Fmoc-protected monomers were also used to

synthesize β-peptoids but led to poor yields. [ 62 ] A ring-opening

polymerization was ? rst investigated by Birkofer et al. with the polymerization of N -p -tolyl-β-alanine- N -carboxyanhydride. [ 63 ] This reaction has been more recently characterized establishing it's living character via kinetic analysis, and the polymerization

has been used to synthesize block copolymers. [ 64 ] However, fur-ther characterization is still required as the molecular weights

cannot be reliably predetermined.

T he single methylene extension of β-peptoids is not the only backbone variant that has been explored. Aminooxy peptoids have been synthesized in both directions (C →N and N →C) using solution phase synthesis combining nitrobenzenesul-fonamide-protected N -substituted aminooxyacetate t ert -butyl

esters (Figure 4b ). [ 65 ] Several pentamers with hydrophobic side

chains were synthesized using this method. Hydrazino azapep-toids have also been synthesized using a solution phase syn-thesis.

[ 66 ] The Kodadek group has additionally worked towards the synthesis of several variants of the typical structure of an α-peptoid. To increase the conformational restraint and func-tionalization of the peptoid, peptoids with side chains on both the α-carbon (peptide-like) and nitrogen (peptoid-like) have been synthesized using a modi? cation of typical submon-omer synthesis. [ 67 ] All of these techniques add functional and structural diversity to the peptoid backbone; however, thus far they have been limited to low molecular weight compounds. Fortunately, the multifunctional submonomers pursued by the Kodadek group are compatible with the typical submonomer technique and could therefore be individually introduced into larger peptoid polymers. T he synthesis of azapeptoids has been achieved by the Kodadek group by using acyl hydrazides in place of the typ-ical primary amine in the submonomer synthesis method.

[ 68 ] The crystal structure of an acylated monomer indicated that acyl hydrazides tend to prefer t rans amide bond geometries. Tetramer azapeptoids were synthesized to demonstrate that short structures could be prepared in high yields; however, this chemistry was found to undergo side-reactions under typical peptoid synthesis conditions as a result of intramolecular cycli-zations. Further work indicated that these side reactions only occur when a methylene group precedes an aromatic group

in the side chains of the azapeptoid.

[ 69 ] To solve this problem, carbazates and semicarbazides were investigated as alternatives to the acyl hydrazides. These did not perform the undesired cyclization when incorporated with typical peptoid monomers. An octamer library combining typical peptoid monomers and those based on acyl hydrazides, carbazates, and semicarbazides was synthesized, demonstrating the compatibility of these reac-tions. There are many commercially available carbazates and semicarbazides, making this a promising technique for incor-porating backbone diversity. Peptoid hydrazides, which have a similar structure to azapeptoids, have been synthesized using

hydrazine as a monomer in the submonomer synthesis. [ 70 ]

After the cleavage of the peptoid from the resin, modi? cation with an aldehyde and reduction with sodium cyanoborohydride yielded a peptoid hydrazide backbone. This chemistry was used with akyl, aryl, and heterocyclic groups functionalizing the aldehyde.

T he Kodadek group has also expanded the variety of peptoid backbone structures by incorporating cyclic molecules into the backbone of peptoids. This backbone design leads to unique structures resulting from the additional conformational con-straints (Figure 4 b ). This conformational restraint is particu-larly advantageous as typical peptoid polymers are much more ? exible than their corresponding peptides due to the loss of the

chiral center. [ 71,72 ] The incorporation of heterocycles including

oxazole, thiazole, and pyrazine structures into the middle of the backbone was optimized using microwave-assisted condi-tions. [ 73 ] These structures did not prevent sequencing with tandem-mass spectrometry of a combinatorial library. In addi-tional work, 2-oxopiperazine was incorporated in the center

of a peptoid backbone.

[ 74 ] Unlike the heterocycles, which were added as monomers, the 2-oxopiperazine was formed on the solid-support by doing an intermolecular cyclization with an amine side chain and a C-terminal 2-chloroacetamide. The pip-erazine ring itself was functionalized with alkyl and benzyl side chains. T he primary structure of peptoids has additionally been modi? ed by creating a variety of nanostructures using branching architectures. A one-pot synthesis was used to

create peptide-peptoid podands built from tripodal skeletons.

[ 75 ] Tri-acid, tri-amine, and tri-cyanate scaffolds were used to syn-thesize a variety of chimeric podands with functional groups including aromatic and heterocyclic rings as well as more com-plex structures such as steroids (Figure 4 c ). The modularity of the synthesis of these podands and cage structures allows for tunable and versatile structures. For instance, the environment formed by a tri-choline compound yields a unique exterior that changes hydrophobicity depending on the solvent microenvi-ronment. These structures have potential applications the rapid identi? cation of new catalysts and receptors. M ultiple groups have explored peptoids as the basis for dendrimers (Figure 4 c ). Bradley and coworkers used N -(6-ami-nohexyl)glycine as the monomer unit of their dendrimer,

which was assembled on a solid-phase resin.

[76]Microwave irradiation was used to accelerate the coupling reactions of the aminohexylglycine monomer. Three generations of den-drimers were synthesized; double couplings were required to ensure high yield in the second and third generation struc-tures. Mass-spectrometry and HPLC were used to analyze the dendrimer products after ether precipitation, demonstrating an 85% yield for the ? nal generation three dendrimer. These structures were evaluated as transfection agents and found to be non-toxic and able to transfect HEK293T cells. More recently, click chemistry has been used to assemble peptoid dendrimers on a hexaphenylxylylene (HPX) and tetraphenyl-methane (TPM) core. [77] Rigid cores were constructed func-tionalized with alkyne or azide moieties, and peptoid arms

were synthesized on solid phase with an N-terminal azide or alkyne-containing monomer and aryl or methoxyethyl side chains. The dendrimers were assembled using Cu-catalyzed

REVIEW

click chemistry and washed with a solution of ethylenediami-netetraacetic acid. Dendrimers with the HPX core were found to be insoluble in organic solvents; however, dendrimers with the TPM core were soluble in polar solvents such as acetoni-trile allowing HPLC puri? cation. The dendrimers were char-acterized using electrospray ionization (ESI) MS and NMR diffusion-ordered spectroscopy (DOSY), allowing the calcu-lation of the hydrodynamic radii of the compounds. All of these approaches to synthesizing branched peptoid structures have explored just a few of the potential structures available to N -substituted glycine oligomers; these structures could be uniquely useful as nanoarchitectures with more tunability and stability than their peptide counterparts.

2.3. C ontrolling the Secondary Structure

A lthough peptides primarily favor a t rans amide bond, peptoid monomers can favor either the c is or t rans orientation of the amide bond depending on local intramolecular interactions ( F igure 5 a ). Therefore, many researchers have explored how to control the conformation of the amide bond by varying the size and electronic structure of the side chains and the interactions between the side chains and the backbone. One of the ? rst investigations of the c is versus t rans state was performed by

Rabenstein et al. [ 72 ] They investigated the kinetics of the isomer-ization of the amide bond with butyl, benzyl, and methyl side

chains in solution using NMR. All of the conformational iso-mers associated with structures containing one, two, and three amide bonds (representing two, four and eight conformations, respectively) were evaluated. The resonances of each isomer were assigned using total correlated spectroscopy (TOCSY) and rotating frame Overhauser effect spectroscopy (ROESY), and

inversion-magnetization transfer NMR experiments were used to study the kinetics of exchange. Interestingly, it was observed that the isomerization rate was dependent on the location of the amide bond within the peptoid, and not necessarily just the side group. On average, 35% of the amides were in the c is conformation, which is slightly higher than the 10% typically observed with proline, the only proteinogenic N -substituted

natural amino acid. [ 78 ] Additionally, the amide bond isomer

tended to be affected by the orientation of neighboring amide bonds: when one amide bond favored a t rans conformation, the neighboring amide bond demonstrated an increased preference for the t rans conformation as well. T he control of c is versus t rans amide bond using side chain electronic effects has been investigated in detail. Loca et al. determined that the n →π * interactions that have been charac-terized in polyproline sequences [ 79,80 ] can stabilize amide bond

conformations in peptoids as well. [ 81,82 ] The n →π *interaction

between neighboring amides serves to stabilize the t rans con-formation (Figure 5 c ), while the n →π * interaction between the amide and aromatic side chains or carbonyl-containing side chains serves to stabilize the c is conformation. The unusual c is conformation was observed in 89% of the amides using a posi-tively charged α-chiral methylpyridine-containing side chain

(Figure 5 b ). Solution phase studies were completed using 1H -NMR and NOESY , and solid-state information was obtained from crystal structures. Gorske and coworkers continued to characterize the n →π * stabilization and found evidence of a

“bridged” interaction. [ 83 ] In this work over 90% of monomers

with an α-chiral naphthyl or pyridine monomer formed a c is amide conformation. For the naphthyl compound, there is

likely a steric component,

[ 84 ] but generally computational anal-ysis determined that for electron de? cient aromatic side chains,

a direct n →π * interaction between the amide and the ring was

the driving force stabilizing the c is amide bond. However, for rings with a higher electron density that contained an α-methyl group, a “bridged” interaction could contribute to the stabiliza-tion. For these cases, the σ * orbital of the methyl group pro-vides a bridge for the electron density from the carbonyl to con-tribute to the π-system of the aromatic ring. K irshenbaum and coworkers performed computational, NMR, and X-ray crystallography studies on N -aryl side chains to determine the conformational preferences of ten aniline sub-stituents of varying size and electronic properties.

[ 85 ]Quantum mechanical calculations on N -methylacetanilide-based mono-mers (the smallest aryl monomers examined) indicated a sig-ni? cant preference for the t rans geometry (Figure 5 c ), and electron-donating groups tended to increase this preference with the reverse trend for electron-withdrawing groups. These results were con? rmed by NMR studies; however, even an aniline with a ? uorine substituent maintained a t rans :c is ratio greater than 9:1. M oieties that lead to an exclusive conformation of the amide bond are rare. Structures have been inserted to create turns in

the middle of the peptoid backbone,

[ 86 ] but only a few functional groups will lead to a single conformation. The Taillefumier group demonstrated that the t ert -butyl side chain (Figure 5 b ) exerts a purely steric effect, forcing the c is conformation of the amide in both water and organic solvents as demonstrated by 1 H -NMR and NOE experiments where a t rans conformation was

F igure 5. S tructural features that induce c is vs. t rans amide bonds. a) The equilibrium between the c is and t rans amide bond is depicted with blue circles highlighting the difference in location of the α-carbons with respect to the plane of the amide bond. b) Turn and side chain moieties that will induce c is amide bonds. c) Structural features that lead to t rans amide bonds.

R E V I E W

undetectable. [ 87 ]

Of particular interest is the research displaying the conformational control provided by the triazole functional

group in the side chains of α and β-peptoids. [ 88 ] In this work,

α and β-peptoids were synthesized with propargylamine and later functionalized with Cu-catalyzed click chemistry. This chemistry has been used for post-synthesis functionalization

in other reports; [ 30,77 ] however, this work demonstrated how the

electron-de? cient triazole, even with additional functionaliza-tion, can control the amide isomerism. A triazolium linkage led to greater than or equal to 89% of the population with c is con-formations. With the neutral triazole, the c is conformation was still preferred, but to a lesser extent. These measurements were completed with NOESY , and the effect was investigated with a computational study of the potential energy surface. Their work attributes the stabilization of the c is amide to both the n →π * interaction and an intramolecular H-bond between the triazole hydrogen and the neighboring carbonyl that is not involved in the n →π *interaction. S econdary structure in a peptoid polymer was ? rst pre-dicted with computational studies of peptoids with chiral side

chains.

[89] This analysis predicted asymmetric Ramachan-dran-like plots and therefore indicated the ability of side chains to induce a conformational preference for the amide bond. Barron and coworkers completed some of the initial analysis on forming peptoid helices using chiral aromatic

side chains.

[90] Since the peptoid structure lacks the amide hydrogen involved in stabilizing peptide helices, other stabi-lizing interactions are required. However, the incorporation of chiral aromatic monomers in excess of 50% of the total number of side chains was suf? cient to confer helicity. Signi? -cantly, a chiral aromatic monomers at the C-terminus seemed to be required, though all of these requirements were relaxed with increasing length of the polymer. These helices were able to withstand temperatures up to 75 °C. A continuation of this work displayed the ? rst crystal structure of a peptoid sequence ( F igure 6 a ) which was a left-handed helix composed of chiral aliphatic side chains, demonstrating that aromatic monomers

were not required for helix formation.

[91] The crystal structure indicated an approximate pitch of 6.7 ? and three monomers per turn, results that are comparable to the polyproline type I helix. Both of these studies were important preliminary dem-onstrations of the capability of secondary structure within pep-toid polymers. More recent work veri? ed the signi? cance of a chiral aromatic monomer at the C-terminus demonstrated by Barron et al. and indicated that in addition to the placement of a chiral monomer at the C-terminus, an α-chiral monomer at the second monomer from the N -terminus seems to also have a strong impact on the secondary structure of the peptoid

polymer.

[92] The presence of only one alpha chiral monomer in a heptameric peptoid placed in that position led to helical character measurable by circular dichroism (CD) spectroscopy. This work highlights the signi? cance of the monomers, and their location in the primary sequence, in directing the peptoid secondary structure. Helices are a fundamental component of protein secondary structure, and therefore the development of peptoid helices is a stepping stone to assembling higher-order structures. Understanding how to control secondary struc-ture formation is critical to the creation of functional peptoid polymer materials.

R esearch on peptoid helices has been expanded by incorpo-rating functionalized chiral aromatic monomers. The synthesis of α-chiral phenyl monomers with p ara -substituted thiols, car-boxylic acids, and amides was completed, and it was demon-strated that these monomers still directed helix-formation. [ 93 ] Peptoid sequences with a α-chiral penta? uorophenyl mono-mers were used to speci? cally probe helix-stabilizing inter-actions. [ 94 ] It was determined using NMR and CD that the hydrogen bonding of the monomer to neighboring heteroatoms was the most signi? cant stabilizing force, and that there was no signi? cant contribution from π-stacking. The placement of the α-chiral penta? uorophenyl monomer at the N-terminus of a nonamer predominantly led to the formation of a “threaded

loop” that was previously identi? ed by Radhakrishnan et al. [ 95 ]

In a similar evaluation of a p ara -nitrophenyl α-chiral monomer, this monomer at the N-terminus also stabilized the formation

F igure 6. U sing the primary sequence to control secondary structure.

a) A list of primary structure characteristics that encourage the forma-tion of a helix. The structure of a helix as generated using molecular mechanics is shown. b) A “threaded loop” secondary structure observed with peptoid nonamers. c) A sheet structure has been observed within peptoid nanosheets. The structure predicted with coarse-grained mod-eling is shown. d) The structure of a unique ribbon secondary structure is shown as an overlay of 10 low-energy structures, as determined by NMR

spectroscopy. a) and b) Reproduced with permission.

[ 96 ]Copyright 2009, American Chemical Society.

REVIEW

of a threaded loop (Figure 6 b ). [ 96 ]

However, the o rtho -nitro-phenyl had a destabilizing effect on the secondary structure of the polymer. Even this small chemical difference between the o rtho and p ara nitro-groups has a signi? cant impact on the folding of the peptoid. T he other canonical secondary structure for peptides, β-sheets, has been observed with peptoids as well. The Black-well group has crystalized small peptoids with N -hydroxyamide peptoid dimers that appear to have β-sheet-like hydrogen

bonding interactions between chains. [ 97 ] While these are very

small structures, it is promising that the hydroxyl side chains could promote β-sheet-like structure in longer polymers. There has been evidence of β-sheet-like structure in longer polymers within the 2D-crystals, or nanosheets, designed by Zuckermann and coworkers (Figure 6c ). [ 98 ] In these assemblies, the inteter-molecular interactions are hydrophobic forces instead of the typical hydrogen bonding, but powder x-ray diffraction (XRD) has indicated a spacing of 4.5 ? between the polymers, indi-cating similar packing to peptide beta-sheets that have spacing

of 4.7 ? due to hydrogen bonding.

[ 99 ] A unique secondary structure rationally designed by Black-well and coworkers is the peptoid ribbon (Figure 6 d ).

[ 100 ]This structure was obtained using alternating bulky α-chiral naph-thyl monomers and achiral aromatic monomers, which led to alternating c is and t rans orientations of the amide bonds. The ribbon itself had a helical conformation that was parallel to the backbone of the peptoid as characterized by X-ray crystal-lography, NMR, and CD. This conformation was maintained in protic and aprotic solvents and at different dilutions of the pep-toid. This work and the previously described research character-izing the secondary structure of peptoid polymers has created a toolbox of moieties that allow for tunable structures while retaining functionality.

2.4. B ulk Properties

W hile precise sequence tunability enables some functionali-ties, an understanding of how to control the bulk properties of peptoid polymers is also important in the development of functional materials. Physical properties controlled by polymer composition include the crystallization and melting transitions. Side chains have been known to impact these properties for over 50 years, and therefore multiple groups have analyzed the impact of peptoid side chains on these properties. An evaluation of poly( N -ethylglycine) with differential scanning calorimetry (DSC) revealed a single glass transition temperature around

100 °C, indicating an amorphous structure.

[ 101 ] This work also evaluated the melting behavior of peptoids with N - propyl, N -butyl, and N - p entyl side-chains ? nding that the melting tem-peratures decreased with the chain length and increased with increasing degrees of polymerization. The Zhang group built on this work by evaluating peptoid sequences synthesized by solution-phase polymerization and with linear alkyl side chains ( F igure 7a ). [ 102 ] Longer side chains led to two melting and crys-tallization temperatures and a loss of the glass transitions, likely due to the decrease in the volume of amorphous regions. The ? rst melting temperature increases with the length of the side chains and was therefore attributed to side chain crystalli-

zation. The second melting temperature and associated enthal-pies decreased with the increasing length of the side chains (for butyl and longer), and was thus attributed to main chain crys-tallization, which is hampered by increased ordering of the side chains. An investigation of branched side chains revealed they prevented packing of both the side chains and the main chain. Wide-angle X-ray diffraction indicated that heating followed by cooling slowly stimulated the crystallization of the peptoid polymers. T o investigate the impact of introducing slight sequence vari-ations on melting temperatures, sequence “defects” have been

introduced to peptoid polymers.

[ 103 ] In work by the Segalman and coworkers, several 15-mer peptoids were synthesized with the primary monomer having either isobutyl or 2-pheny-lethyl side chains with either two or four “defect” monomers (2-methoxyethyl, hexyl, or 3-phenylpropyl). The introduction of 2-methoxyethyl and hexyl defects into the isobutyl 15-mers

F igure 7. C haracterized bulk properties of peptoid-based homopolymers.

a) Melting transition temperatures of the main chain and side chain groups of alkyl homopolymers as measured with differential scanning calorimetry (DSC). An example DSC thermogram is shown for a hexyl homopolymer showing both melting transitions. b) Various peptoid poly-mers with small alkyl side chains were evaluated for thermoresponsive properties. Three were identi? ed to have lower critical solution tempera-tures (LCSTs) in water within a physiological range. c) The conductivi-ties of 20-mer polyethylene oxide mimetic polymer electrolytes were measured at a variety of lithium salt concentrations and temperatures to elucidate the structure-conductivity relationships of sequence-de? ned

polymers. a) Reproduced with permission.

[ 102 ] Copyright 2013, American Chemical Society.

R E V I E W

led to a slight change in packing as indicated by XRD, but a signi? cant decrease in the melting temperature (up to 20 °C) with a continued decrease with increasing number of defects. Interestingly, the location of the defects also has an impact on the melting temperature and a slight impact on the packing. With precise control over the locations of the defects, it was determined that substitutions towards the center yielded the lowest enthalpies of melting, whereas substitutions near the termini yielded the highest enthalpies of melting. Surprisingly, the 3-phenylpropyl defects in the 2-phenethyl peptoid polymer completely prevented crystallization. S imilar classes of peptoid polymers have been evaluated for their thermoresponsive properties in water (Figure 7 b ). Copoly-mers of methyl and butyl monomers were discussed earlier, [52] but copolymers constructed from butyl- and ethyl-containing monomers were also evaluated by measuring the UV-vis trans-mittance over a range of temperatures. [104] The cloud point

temperature could be tuned between 20–60 °C by varying the ratio of each monomer. Homopolymers with n -propyl,allyl,and isopropyl were also determined to have cloud point tem-peratures within that range (15–25 °C, 27–54 °C, and 47–58 °C

respectively).

[105] Due to semi-crystallinity of the propyl and allyl substituted samples, they had to be treated with methanol in order to solubilize them in water. Notably, propargyl sub-stituted monomers resulted in a polymer that was insoluble in water. The precipitates formed by polymers of the propyl and allyl monomers above the cloud point formed a rose bud architecture as characterized by SEM. Polypeptoids with allyl side chains have been further studied to determine their ther-moresponsive behavior and their behavior upon further photo-chemical modi? cation with thiol containing moieties.

[106] H omopolymers of monomers bearing ethylene oxide-based substituents have been evaluated as polymer electrolytes for battery applications due to the high conducting capacity of poly(ethylene oxide) polymers containing dissolved lithium salts (Figure 7c ).[107] The polymers were synthesized using step-wise solid-phase submonomer synthesis, allowing for the study of discrete polymer chains. Characterization was per-formed using small-angle X-ray scattering (SAXS), XRD, and DSC to determine that the structures are non-crystalline and have glass-transition temperatures, which vary based on the monomer. The transition temperature is increased with the introduction of lithium salts and increases linearly with con-centration. The maximum conductivity was obtained with the polymer with the longest ethyleneoxy side chain, though this conductivity was dependent on the ion concentration. Increasing concentrations of lithium salts leads to increased conductivity; however, eventually the decreased chain mobility becomes a more signi? cant effect and the conductivity starts decreasing with increased salt concentrations. The highest

conductivity obtained (2.6 × 10

?4 S/cm) was higher than those obtained with peptide-PEO mimetics. Interestingly, these block copolymers exhibited lamellar nanoscale phase-separated

morphologies over a wide range of compositions.

[108] It was also shown that the triethyleneoxy-containing peptoid block is amorphous as a homopolymer, but it will crystallize when appended to an N -decyl block ( F igure 8a ).[109] This enabled a direct comparison of the conductivities of an amorphous

and crystalline phase of the same peptoid block. [110] The tun-

ability of these structures will allow for future detailed anal-ysis of modulation of ion conductivity by varying the primary structure. B lock copolymers and related materials are a natural research direction for peptoid polymers, but one that is only

recently being extensively pursued. Polymerization,

[111]step-wise synthesis, [108] and a combination of both [112] have been

used to synthesize block-copolymers. Some of the ? rst work on block-copolymers used peptoid polymers and polystyrene linked via click chemistry to study how to the peptoid composi-tion affected the solid-state polymer assembly (Figure 8 b ). [112]

An increase in the glass transition temperature was noted with increase in the ratio of aromatic monomers to 2-meth-oxyethyl monomers. Block copolymers of peptoids with a

F igure 8. I nvestigations into the morphologies formed by block copoly-mers of peptoids (spheres) and polystyrene (solid lines). a) The mor-phologies of a completely peptoid-based block copolymer were evaluated

with 38-mer diblocks. The polymer with the linear aliphatic substituents formed a crystalline block (TEM shown), while the shorter branched block formed amorphous structures conjugated to the same conducting block. b) Schematic of structures leading to lamellar and hexagonally packed morphologies based purely on the composition of the peptoid segment. Ratios between 4:6 and 6:4 styrene:methoxyethyl led to lamellar mor-phologies while 7:3 and 8:2 led to hexagonally packed structures. c) A block copolymer composed of three different monomers was used to evaluate how surface characteristics were affected by the placement of a ? uorinated block.

REVIEW

block of N-(2-methoxyethyl)glycine and a polystyrene block assembled into lamellar and hexagonally packed morpholo-gies, as indicated by transmission electron microscopy (TEM) and SAXS analysis. As expected, composition ratios close to 1:1 form lamellar morphologies, and ratios closer to 7:3 or 8:2 form hexagonal morphologies. With the incorporation of N-2-phenylethyl side chains into the peptoid block, the phenyl monomers started mixing with the polystyrene phase, leading to decreased order in the assembled structure. This demon-strated that the modular sequence of the peptoid-block ena-bles tuning of the segregation strength in the three-dimen-sional structure. T he ability to synthesize discrete molecules as blocks of multi-block copolymers has enabled detailed studies of surface exposed monomers in the assembly of a multi-block lamellar

structure (Figure 8 c ). [113] These polymers were synthesized

for antifouling applications, but provide interesting insight into the blurred borders between the lamellae of assembled block copolymers. Segalman and coworkers used a triblock copoly m er, containing blocks of various lengths composed of benzyl, 2-methoxyethyl, and ? uoroalkyl monomers, to explore the effect of the monomer ratios on the moieties exposed to the surface. X-ray absorption ? ne structure spectroscopy (NEXAFS) was used to characterize the exposed surface of the block copolymers. The surface segregation increases with the number of ? uorinated monomers incorporated. Although lamellae are commonly described as having a de? ned boundary, there is typically signi? cant mixing between the layers, and a block of ? ve-? uorinated monomers increases this mixing such that peptoids up to 30-mer in size are unable to completely cover a surface from polystyrene exposure. Shifting the location of the ? uorinated block to the middle of the pep-toid structure decreases the thickness of the peptoid layer, increasing the amount of polystyrene displayed at the surface.

3. S urface Immobilized Polymers and Coatings

3.1. A ntifouling and Antimicrobial Coatings

O ne important application of sequence speci? c polymers is to create functional coatings on materials. Maintaining the integ-rity of surfaces, from instruments as complicated as medical devices to surfaces as large as a ship’s hull, has recently become

a focus of ongoing research.

[ 114 ] Coatings that can prevent the adhesion of proteins, bacteria, and marine microbes are incred-ibly important in maintaining the useful lifetime of materials. The Messersmith group began the application of polypeptoids

as antifouling materials.

[ 37 ] To achieve this, they focused on a class of peptoid polymers with short, hydrophilic side chains. They used a 25-mer of N -(2-methoxyethyl) glycine with a block of ? ve monomers designed to mimic mussel adhesion peptides containing L-3,4-dihydroxyphenylalanine (DOPA) and basic pri-mary amines. The 2-methoxyethyl monomers were chosen to enhance antifouling behavior – namely water solubility, charge neutrality, and the ability to accept hydrogen bonds. Over a period of ? ve months, titanium-modi? ed surfaces modi? ed with these peptoid structures resisted cell adsorption as meas-ured by ? uorescence microscopy.

T his work was recently expanded by a series of experimental

and theoretical analyses. The same anchoring pentamer was used to immobilize varying lengths of 2-methoxyethyl pep-toid polymers, [ 115 ] monodisperse polysarcosine sequences, [ 116 ]

homopolymers of hydroxyl-containing monomers, [ 117 ] and het-eropolymers with alternating charged monomers and 2-methy-oxyethyl monomers. [ 118 ] Taking advantage of the tunability of

the peptoid structures, zwitterionic heteropolymers of different lengths and different spacings between the charged monomers were synthesized and evaluated for antifouling capabilities. Slight differences were observed in the adherence of different cell types, which could be attributed to the aforementioned moieties. In another variation of this work, click chemistry was used to attach sugar units to the previously evaluated meth-oxyethyl peptoid polymer linked to the anchoring pentamer ( F igure 9a ). [ 119 ] These glycocalyx mimicking structures ef? -ciently prevented a series of cells and proteins from adhering to a titanium surface, as demonstrated by ? uorescence microscopy (Figure 9 b ). The results of molecular dynamics simulations indicate that sterics are likely the largest contributing factor to the antifouling behavior, in addition to water molecules bound to the terminal sugars. In a unique application of antifouling coatings, the methoxyethyl peptoid polymer and pentamer anchor were used to prevent protein adhesion on the imaging

F igure 9. C hemical structures and an example application of antifouling

peptoids. a) Peptoid structure terminating in a peptide mimetic of a DOPA-containing mussel inspired adhesion peptide linked using click chemistry to carbohydrates as a glycocalyx mimetic. b) Representative images of titanium surfaces with and without the peptoid-carbohydrate antifouling coating treated with ? broblasts demonstrate the ability of the glycocalyx mimic to prevent cell adhesion. Reproduced with permis-sion.

[ 119 ] Copyright 2013, American Chemical Society.

R E V I E W

surface used for single-molecule microscopy. [ 120 ]

Currently poly(ethylene glycol) methyl ether is used for this application, but this coating is labile to hydrolysis and can have unpredict-able grafting densities. The peptoid polymer was demonstrated to be more effective at blocking non-speci? c protein interac-tions than the typical coating, demonstrating the practicality for future single molecule studies. P eptoid polymers synthesized with living polymerizations have also been used as antifouling coatings. Poly(β-peptoids) were synthesized using cobalt catalyzed ring-opening poly-merization and adhered to a gold surface using a terminal

thiol. [ 57 ] Surface plasmon resonance spectroscopy was used to characterize the degree of protein adsorption. This was dem-onstrated to be an effective and synthetically scalable platform for coating surfaces. Ring-opening polymerization performed directly from the surface has also been applied and provides

greater surface polymer densities. [ 53 ] An amine-silane was used

to functionalize a glass surface, and block copolymers were syn-thesized directly from the amine moiety. Atomic force micros-copy scans of the surface indicated a smooth surface with less than a 1 nm average roughness. Amphiphilic brushes were synthesized using two sequential surface initiated polymeri-zations and characterized using FTIR spectroscopy and X-ray photoelectron spectroscopy (XPS). A nother approach to anti-fouling is the immobilization of antimicrobial materials on the surface. This approach has been

frequently applied with antimicrobial peptides.

[ 121 ] The Barron group has recently published a number of studies mimicking

antimicrobial peptides,

[ 122,123 ] and the Winter group has applied combinatorial chemistry to this problem.

[ 36,123 ] However, there is only one report of antimicrobial peptoids being immobilized

on a surface.

[ 124 ] In this work, the immobilized antimicrobial peptoids unfortunately did not prevent adhesion of bacteria to the surface, but the immobilized structures were deter-mined to kill the bacteria that adhered. Although more bacte-rial adhesion was measured on a surface coated with polymers of methoxyethyl monomers and the antimicrobial peptoids than on a surface coated with the N -(2-methyoxyethyl)glycine polymers alone, the antimicrobial retained bactericidal effects. Future studies could involve a more detailed investigation of the ratios and lengths of the polymers, in addition to evaluating other antimicrobial structures.

3.2. C hromatography Supports

T he ability to separate enantiomers of natural products and syn-thetic drugs is critical in a wide variety of biomedical applica-tions. A variety of solid supports modi? ed with ligands ranging from small molecules such as amino acids and cyclodextrins to

full proteins have been developed to address this concern.

[ 125 ] Their stability, ability to from secondary structures, and variety of available chiral submonomers make peptoids uniquely useful for this application. The Liang group has published a series of articles evaluating silica modi? ed peptoid structures for these separations. In preliminary studies, polymers of N -( S )-1-phenylethyl)glycine monomers ranging from three to seven monomers in length were evaluated for their ability to isolate enantiomers of twelve different BINOL (1,1′-bi-2-naph-thol) derivatives. [ 126 ] A dramatic increase in separation ability

was seen with small increases in chain length. For BINOL itself, increasing the chain length signi? cantly enhanced the separation ability, but this reached a maximum at 6 monomers, after which the separation ability started to decline. Methylation of the hydroxyls eliminated the separation ability, and placing bulky groups near the hydroxyls reduced the separation ability, con? rming that hydrogen bonding between the substrates and the solid-phase serves as the primary interaction. F uture studies systematically varied the peptoid backbone to assess which other factors contributed to the ability to per-form chiral separations. A series of peptoids built on a trimer of ( R )- N -(1-phenylethyl)glycine evaluated the effect of varying the N-terminal moieties ( F igure 10 a ). Various branched and unbranched alkyl chains, in addition to cyclic chiral and aro-matic monomers, were evaluated. The chromatographic sup-ports with achiral monomers at the N-terminus outperformed the polymers of ( R )- N -(1-phenylethyl)glycine monomers of the same length. The authors hypothesized that this was due to the removal of unnecessary π-stacking interactions with the aromatic analytes. Additionally, the introduction of additional chiral structures to the N-terminus such as N ′-phenyl -L -proline and N ′-phenyl -L -leucine led to increased peak separation for speci? c analytes. Another study varied the chirality of the trimer ( R )- N -(1-phenylethyl)glycine monomers and determined that heterogeneous chirality unexpectedly led to increased sepa-ration ability.

[ 127 ] T he study of peptoid-based moieties for the separation of chiral products was expanded by creating a variety of branched structures. In recent work, a series of six branched structures

F igure 10. P eptoid structures analyzed for their performance in chiral separations. a) The effect of the N-terminal monomer of a trimer of ( R )- N -(1-phenylethyl)glycine groups was investigated. b) The modi? ca-tion of branched peptoid structures with quinine and quinidine moieties was also evaluated.

REVIEW

with six chiral centers was evaluated as chromatography sup-ports; again the chirality was demonstrated to have a large

effect on separation ability. [ 128 ] Heterogeneity in the chirality

of neighboring structures led to increased separation ability, with one structure performing better than some commercial chiral chromatography columns. To increase the ef? cacy and analyte compatibility, the peptoid-base was expanded by modi-? cation with quinine and quinidine moieties, which have been demonstrated to be broadly applicable in chiral separations (Figure 10 b ). [ 129 ] These chromatography supports were dem-onstrated to outperform commercial quinolone-based supports for a variety of analytes. In some separations, the quinolone or quinidine moieties dominated the interactions with the analyte; however, for some of the separations hydrogen bonding and π -stacking with the peptoid played the most signi? cant role in chiral recognition. The series of reports of peptoid-based chro-matography supports demonstrate not only the applicability of peptoids for this application, but additionally how the sequence modularity can be used to create a library of supports tunable for the analytes at hand.

3.3. M aterials Developed with Combinatorial Libraries

C ombinatorial chemistry has played an integral part in indus-trial drug discovery in recent years, and thus a variety of tech-niques for spatially resolved immobilization of peptoids have been explored. One of the ? rst was an adaption of the SPOT

synthesis initially developed for peptides.

[ 130,131 ] In this method, step-wise synthesis is performed on a cellulose membrane

sheet where each individual reaction is spatially separated from its neighbor, leading to a library of thousands of compounds in a 2D array on the membrane. This technique was applied to the synthesis of a peptoid library in which 8000 peptoid hexamers were synthesized using the a modi? cation of the submonomer

method. [ 43 ] Due to the prevalence of free hydroxyl groups on

the surface of the cellulose, bromoacetic acid 2,4-dinitrophenyl-ester was used for the acylation step. The reliability of the syn-thesis technique was veri? ed by cleaving individual peptoids and quantifying the purity with LC-MS (on average 44% purity), and the screening capabilities were evaluated by screening the library for antibody af? nity. A more common method of creating a spatially separated array of unique peptoids structures is to create microarrays on functionalized glass. The construction of microarrays from small molecules to proteins has been approached from a variety of perspectives, and there has been signi? cant progress in

their construction in recent years. [ 132,133 ] A microarray com-posed of a peptoid library was developed that created unique

patterns for individual proteins, demonstrating the promise of

peptoid libraries in biomedical detection devices.

[ 134 ] For high yielding and pure peptoids, the synthesis was completed on a styrene-resin and then adhered to the maleimide-functional-ized glass using a SpotArray 72 Microarray Printing System. Using microarrays of over 7500 compounds, they were able to create a “? ngerprint” of hybridization unique to individual representative proteins. The protein was either ? uorescently labeled or visualized with a ? uorescent secondary antibody, and the patterns identi? ed were both unique to the proteins of interest and reproducible ( F igure 11 a ). This method could

F igure 11. A pplication of combinatorial chemistry to the development of materials. a) Unique protein ? ngerprints. Peptoid microarrays displaying

7,680 distinct compounds (library schematic shown) were incubated with ? uorescently labeled ubiquitin (Ub) and maltose binding protein (MBP). The ? uorescence intensities were assigned to a color coding scheme as shown on the right, and depicted as a bar code for ease of visualization of binding. The expanded portion further illuminates the differences in the binding patterns. b) Metal binding peptoids identi? ed using combinatorial chemistry were immobilized on polystyrene solid-supports (grey sphere). The ability of these structures to remove chromium(VI) selectively, a heavy metal contaminant, was evaluated by exposing the immobilized peptoids to various solutions with simulated chromium(VI) contamination. a) Reproduced

with permission.

[ 134 ] Copyright 2005, PNAS.

R E V I E W

reliably detect less than 100 nM of a protein in the presence of E . coli cell lysate by subtracting background signal from the lysate proteins (and, when relevant, the secondary antibody). This detection limit is well below the binding af? nities previ-ously identi? ed for peptoid-protein interactions, indicating the array itself is critical for the detection of low concentrations of proteins.

I n variations of this work, the Kodadek group has also syn-thesized arrays of unique cyclic peptoids [ 135 ] and applied micro-arrays in the identi? cation of high af? nity ligands for proteins

of interest. [ 44,136,137 ] The immobilization of cyclic peptoids was

completed using maleimide-functionalized glass reacted with a cysteine adjacent to the peptoid sequence. The peptoid was cyclized following the on-bead deprotection of a carboxylic acid, which was then reacted with the N-terminus. After cleavage from the resin, the cysteine was robotically spotted onto the glass. A particularly interesting application of the microarrays in the development of diagnostics is the identi? cation of ? n-gerprint patterns for antibodies characteristic of Alzheimer’s

disease.

[ 44 ] In this study, blood samples from patients were used to develop diagnostic readouts from a microarray of approxi-mately 15 000 peptoid octamers. A nother unique application that demonstrates the applica-bility of combinatorial chemistry for materials applications is the identi? cation of peptoid ligands. Some of the ? rst work towards the identi? cation of metal-binding peptoids was com-pleted by Kirshenbaum and coworkers.

[138] The binding of a pentamer and hexamer with terminal 8-hydroxy-2-quinoline

methylamine monomers, and the af? nities for Cu

2+ and Co 2+,were characterized using titrations monitored with UV-vis and CD spectroscopy. The development of peptoid-based ligands has been continued with the development of cyclic peptoids

which coordinate Na

+ and Gd 3+,[139,140] and combinatorial assays have been established to identify ligands with selec-tive af? nity for individual metal ions in a heterogeneous mix-ture. [38,141] X-ray ? uorescence has been used as a screening technique for the identi? cation of multiple metals coordinated

to an individual resin-immobilized sequence.

[141] As a proof-of-concept, this technique has been used to identify ligands

for Ni

2+ that can remove these ions from a buffered solution. Francis et al. used a colorimetric assay, based on previous work

with peptide-based ligands, [142,143] to identify ligands which could coordinate toxic Cr 6+ in multiple environments that sim-ulated natural contaminated water (Figure 11 b ).

[38] The ligands identi? ed in the combinatorial screen were able to reduce the concentration of the contaminant to levels comparable to the EPA limit for drinking water. This work also includes struc-tural variants and NMR characterization to illuminate inter-actions contributing to binding and selectivity. These studies demonstrate the capabilities of peptoids as metal ligands and highlight the use of combinatorial chemistry in identifying selective ligands. Though the applications of selective ligands are broad (therapeutics, imaging agents, sequestration, etc.) the factors contributing to selective binding are not well under-stood. [144] Due to the limited precedent for the rational design of selective ligands, these studies demonstrate the value of combinatorial chemistry for rapidly identifying new struc-tures for unique materials properties including selective metal coordination.4. S upramolecular Assemblies

4.1. C ontrolling the Crystallization of Materials

T he crystallinity of biominerals like bone and abalone nacre is precisely controlled in nature by native biopolymers. The effort to reproduce this level of morphological control with inexpen-sive, well-de? ned synthetic materials has been approached both by studying the natural systems and by investigating the

effects of synthetic polymers. [ 145 ] Recently, peptoid polymers

have been applied to controlling the crystallization of calcite

(CaCO 3), [ 146 ] and as antifreeze agents controlling ice crystalliza-tion. [ 147 ] Marine organisms naturally convert CO 2 to carbonate biominerals, and the nucleation of this mineralization has been demonstrated to be dependent on protein and peptide struc-tures. [ 148,149 ] This is also an important topic of research as the sequestration of atmospheric CO 2 becomes increasingly impor-tant with the continued consumption of fossil fuels. P eptoids have been rationally designed to mimic these min-eralization proteins and have been demonstrated to control the

growth of calcite.

[ 146 ] Deyoreo, Zuckermann, and coworkers designed eight structures (examples shown in F igure 12a ) based on the characteristics identi? ed to impact the nucleation of cal-cite (e.g., the balance between hydrophobic and coordinating monomers). An analysis of the morphology of crystals grown in the presence of these peptoids revealed structures similar to those grown in the presence of the native proteins is shown in Figure 12 b . Both the hydrophobic monomers and the number of acid monomers had an impact on the crystal formed, dem-onstrating signi? cant tunability in the crystal morphology. Additionally, the AFM characterization of crystal growth indi-cated that a speci? c peptoid increased the rate of calcite crystal growth by a substantial 20-fold at nanomolar peptoid concentra-tions. Interestingly, at higher peptoid concentrations the degree of rate acceleration lessens, possibly due to surface binding. The authors suggest three possible mechanisms of rate enhancement: (1) coordination of calcium leads to an increased local concentration, (2) coordination decreases the energy bar-rier of desolvation for crystallization, and (3) the water layer is disrupted, decreasing the energetic cost of removing water from the crystal surface.

T his work was continued recently with a series of 28 peptoid chains designed to continue the systematic evaluation of how simple sequence changes can impact the rate of calcite crystal

growth and the ? nal morphology.

[ 150 ] In this study, the focus was on tuning the hydrophobicity of the structures by varying the number of hydrophobic monomers and their substituents. Hydrophobicity of the peptoid was seen to correlate with the reduced formation of {104} faces, but this could be modu-lated by the arrangement of the monomers in the sequence. Increasing electrostatic interactions between the anionic pep-toid polymer and the calcite was seen to increase the formation of {104} faces as well, but the hydrophobicity was demonstrated to have a dominant effect. As the concentration of the peptoids was increased, the formation of {104} faces was decreased due to higher adsorption of the peptoid additives to the calcite sur-face. The ability to exert morphological effects on calcite growth was determined to be an effective method of preliminarily iden-tifying structures that would accelerate the growth of calcite at

REVIEW

low peptoid concentrations. With additional AFM-based char-acterization of calcite-peptoid interactions, a model for pep-toid interaction with the calcite was proposed (Figure 12 c). The authors suggested that their work supports the third mechanistic proposal above: the adhered water is disrupted by peptoid-calcite interactions. The amphiphilic characteristics of the peptoids allow for diblock-like assembly, promoting growth as the carboxylic moieties reversibly interact with the surface of the crystal. The ability to mimic biomineralization is an exciting new application of peptoid polymers, given the signi? -cant amount of control obtained with peptoid polymers in cal-cite crystal growth.

I n a similar application, peptoid polymers were investi-gated for their ability to control ice crystallization as antifreeze agents. [ 147 ]Despite ice being the most prevalent crystalline structure in nature, its crystallization is not fully understood. Antifreeze proteins in ? sh have evolved to control the crystal-lization of ice, arti? cially depressing the freezing point of water. These proteins have been extensively studied and mimicked with synthetic polymers; however, the details of the requisite interactions with ice are not completely understood. [ 151 ]The sequence control of peptoid polymers allows a speci? c inves-tigation of the mechanisms of controlling ice morphologies and crystal growth. Homopolymers of lengths ranging from three to seven monomers of sarcosine, hydroxyl-containing monomers, and methoxyethyl monomers were synthesized, and their impact on ice morphology was evaluated with vari-able temperature video microscopy and compared to glycerol and oligoserine standards. Crystals grown in the presence of any of these additives were smaller than those grown in pure water, but the smaller tripeptoids led to the smallest crystals with the trimer of the hydroxyl-containing peptoid being the most effective. The rate of crystal growth was also impacted, with the trimer and hexamer peptoids displaying the slowest growth rates. Differential scanning calorimetry was used to evaluate the impact of the trimer additives on the melting tem-perature of the ice. The melting temperature was depressed for all of the samples with additives. The peptoid timers and serine trimer displayed a similar depression, with the exception of the sarcosine trimer which had a smaller effect. For each of these additives, the effect was determined to be more signi? cant than that of colligative properties alone. By DSC, the hydroxyl-con-taining peptoid demonstrated the largest temperature change relative to the anticipated colligative effect. Given the successes of the hydroxyl-containing peptoid trimer in minimizing crystal growth and depressing the melting point, it would be inter-esting to continue to evaluate other monodisperse structures to further highlight the functionalities involved in controlling ice crystallization.

P reventing the crystallization of hydrates is also critical in the ?eld of natural gas transportation. [ 152 ]The solid hydrate can form at low temperatures and high pressures, blocking transportation of natural gas. Typically poly(N-methacrylamide) derivatives are used for this application, but the amide back-bone and hydrophobic side chains of peptoids are particularly useful for inhibiting the nucleation of hydrate crystals. [ 153 ]J ia and coworkers evaluated poly( N-alkyl-β-alanine) homopoly-mers and block copolymers as kinetic hydrate inhibitors by

F igure 12. S tudies of the peptoid modulated nucleation and growth of calcite have identi? ed hydrophobicity and the fraction of carboxylic acids as characteristics affecting its growth. Eight structures were designed to evaluate these factors; a) some of these structures (i-iv) are shown. b) Crystal morphologies resulting from these four structures are shown. c) A model for the interaction between a peptoid structure and the calcite {104} face is shown, where the hydrophobic peptoid regions are represented in gray and the hydrophilic regions are in blue. b) Reproduced with permission. [ 146 ] Copyright 2011, American Chemical Society. c) Reproduced with permission. [ 150 ]Copyright 2014, Nature Publishing Group.

R E V I E W

evaluating the solubility at relevant pressures, the cloud points,

and performance of the polymers in simulations of the real system. Polymers with longer akyl side chains performed better than shorter ones, and interestingly a random copolymer per-formed better than a block copolymer of the same composition. While peptoids have only begun to enter into the ? eld of con-trolling crystallization, these studies demonstrate the promise of these highly tunable polymers for this application.

4.2. S olid-State Crystal Packing of Cyclic Peptoids

S ince the ? rst peptoid crystal structure was solved, [ 91 ]there have been a variety of interesting solid-state structures of pep-toids that have been discovered. A recent review extensively covers the structures that have been identi? ed for cyclic pep-toids, [ 17 ] and here we highlight just a few interesting functional structures. In work by De Riccardis and coworkers, two of three synthesized cyclic peptoid structures formed a supramolecular

assembly triggered by coordination of Na +( F igure 13a ). This

assembly is similar to metal organic frameworks (MOFs) and demonstrates the potential application of peptoids as tunable linkers for MOFs. The linkers control the assembly and ? nal

properties of MOFs; [ 154,155 ] thus, the tunability of peptoids is

particularly applicable. Even with the sodium MOFs, slight dif-ferences between the peptoid structures evaluated (Figure 13 a ) lead to different solid-state properties. Of the three structures evaluated, the third structure does not crystallize, and the others lead to different coordination spheres despite containing the same coordinating moieties. O rdered solid-state structures have also been obtained without metal coordination. A cyclic octamer (Figure 13 b ) was designed with alternating c is and t rans bonds to have predict-able placement of the side chains. [ 156 ] The cyclic octamers self-assembled into a nanotube crystal structure with associ-ated water molecules, demonstrating the ability to anticipate secondary structure (Figure 13 c ). The mole percent of water molecules was found to be variable depending on the crys-tallization environment, indicating possible diffusion of the water molecules through the nanotube crystal. Interestingly, the nanotubes were found to be stable up to temperatures of 300 °C, indicating the robustness of this structure. Given the applications that have been identi? ed for peptide-based nano-tubes, including in microelectronics and as biosensors,

[ 157,158 ] it is likely that further research on the stable peptoid nanotubes will reveal additional applications in nanotechnology. O nly one cyclic nonamer (consisting of a repetitive ( R )- N -(1-phenylethyl)glycine peptoid) has been crystallized, but what made the study of this nonamer particularly interesting was the comparison of its crystal structure to the structure predicted by

molecular modeling.

[ 159 ] We will discuss molecular modeling in more detail in a future section, but this highlights how the ? eld is rapidly progressing towards predictable design of peptoid structures in the solid-state.

4.3. S olution-Phase Assembly of De? ned Nanostructures M oving beyond secondary structure, the development of ter-tiary and quaternary structural elements is the next step toward

establishing protein-like control and functionality within pep-toid-based synthetic materials. Supramolecular assembly is an expansive ? eld with building blocks including small molecules, peptoids, polymers, and combinations thereof, and applications range from stimulus responsive materials to semiconducting

nano? bers. [ 160–162 ]As sequence-de? ned polymers, peptoids

could have an unprecedented diversity of conformations with the structural control necessary to take advantage of these. The design of supramolecular assemblies from peptoid polymers has just begun (examples shown in F igure 14a -c), but as more is learned about the properties of peptoids, this could be a rap-idly expanding area of research. O ne approach to the development of supramolecular peptoid assemblies has been to mimick naturally assembling peptides. Two peptoid structures have been synthesized that mimic self-aggregating amylin(20–29) that is associated with type II dia-betes. [ 163 ] One of the peptoids is shown in Figure 14 d and a TEM of the assembly is shown in Figure 14 a . While this struc-ture had some impact on the assembly of amylin(20–29), a pep-toid with the reverse sequence inhibited the formation of pep-tide ? brils for at least four days. The self-assembling ability of these peptoid structures is particularly interesting because they

F igure 13. a ) Structures of three hexameric cyclic peptoids, two of which

form assemblies upon coordination of Na + . b) A cyclic octamer that self-assembles into a nanotube without assistance from metal coordination.

c) Crystal structures of the assembly of the middle structure in a) upon

exposure to Na

+ (left) and of the cyclic octamer in b) (right).

REVIEW

have lost most of the hydrogen bonding pairs of the natural peptide. The assemblies are different than that of the natural peptide, and yet they still self-assemble.

O ne of the most unique structures that has assembled from peptoid polymers is a superhelix (Figure 14 b ) that was formed by an amphiphilic block copolymer. [ 164 ] Superhelices have been constructed from peptide-based copolymers before,

[ 165 ] but this again demonstrates that the capabilities of peptoids are not limited by an achiral backbone. A 30-mer peptoid containing a block of 15 N -(2-phenylethyl)glycine moieties and a block of 15 N -(2-carboxyethyl)glycine moieties was synthesized on solid-phase (Figure 14 e ). After puri? cation the structure was allowed

to assemble in water buffered to pH 6.8. Within 24 hours, nanosheets assembled; after 4–7 days, they transitioned into superhelices (a model of the assembly is shown in Figure 14 f ). Once formed, the superhelices were incredibly stable in solu-tion, lasting for months. The helices were characterized by SEM, TEM, AFM, and X-ray scattering, and were determined to have widths of approximately 625 nm and lengths varying from 2 to 40 μm. In an effort to relate the superhelix chirality to the molecular-scale peptoid chirality, chiral groups were introduced into the hydrophobic block. After alteration of one monomer had no impact on the structure, the entire block was replaced with chiral monomers. However, the superhelix chirality remained the same. This highly ordered structure holds promise for not only mimicking natural molecules that assemble into hierarchical helices, but additionally in gaining a more precise understanding of their self-assembly. A n important assembled structure in bioengineering is the hydrogel, which has applications ranging from sensors to tissue

replacement materials. [ 166 ] Peptoid-based hydrogels are advanta-geous due to their biocompatibility and resistance to proteases.

One approach to the construction of functional hydrogels was to combine a hydrophobic phenyl-based peptoid block with bio-active peptide trimers. [ 167 ] Four peptide sequences were chosen,

and, in combination with the hydrophobic peptoid block, all of them formed hydrogels in phosphate buffered saline buffer (pH 7.4). The gels were characterized by rheometry and TEM to con? rm gel formation and characterize the assembled struc-ture. The oligomers with hydrophilic peptides assembled into sheet-like structures, and those with hydrophobic peptides assembled into networks of ? bers. To con? rm the biocompat-ibility of these gels, various cell lines were treated with the poly-mers in solution and in the gel and their rates of growth were evaluated. Another approach to the development of peptoid-based gels used a small library of peptoid dimers and a variety

of solvents (not just water).

[ 168 ] Optical micrographs of the gels formed show structures varying from uniform ? brous assem-blies to heterogeneous sheets. A nother supramolecular morphology accessible to pep-toid oligomers is microspheres (Figure 14 c ). [ 169,170 ]These have applications analogous to the previously described gel-like structures in drug delivery and sensing and similarly

allow for a systematic evaluation of the molecular interactions leading to self-assembly. A panel of peptoid 12-mers with chiral

helix-inducing, methoxybenzyl, acidic, and basic monomers was evaluated for self-assembly properties (example shown Figure 14 g ). [ 169 ] A comparison of two structures with very similar compositions showed that changing a single monomer

made the difference between microsphere formation and no assembly at all. The microspheres that did assemble also varied

ten-fold in size. This work was continued by investigating the properties required to coat these spheres uniformly on a glass

or silicon substrate. [ 170 ] A microsphere coating could be advan-tageous for allowing ligand binding and for preventing surface

fouling. The administration and drying techniques and solvent were seen to impact the microsphere coating process. Adminis-tration by dipping led to no microsphere coating, while admin-istration of larger volumes followed by evaporation led to the

formation of a well-distributed layer of microspheres. A recently developed controllable class of peptoid supramo-lecular assemblies is two-dimensional crystals, or nanosheets

(

F igure 15). [ 98 ] The combination of two peptoid polymers, one that alternates hydrophobic and basic monomers and another that alternates hydrophobic and acidic monomers, leads to the

formation of sheets with a thickness of 27 ? as indicated by analysis with powder XRD. The assembly takes place at micro-molar concentrations in neutral aqueous solution. Most recent research on nanosheet-like structures has focused on graphene sheets, which may be prepared via exfoliation or organic syn-thesis, because of their potential applications in electronics.

F igure 14. S elf-assembled structures from amphiphilic peptoid poly-mers. a) TEM image of nanoribbons formed by the assembly of the pep-toid structure in panel d. b) AFM image of a superhelix formed from the amphiphilic peptoid block copolymer shown in e). A model of the assembly is shown in panel f. c) SEM image of microspheres formed from structures in panel g. These images demonstrate the size diversity

afforded by a simple rearrangement of monomers. a) Reproduced with

permission.

[ 163 ] Copyright 2007, Elsevier. b) and f) Reproduced with per-mission.

[ 164 ] Copyright 2010, American Chemical Society. c) Reproduced with permission. [ 169 ] Copyright 2013, RSC Publishing.

R E V I E W

However, two-dimensional crystalline materials also have potential applications in sensing, separations, device fabrica-tion, and as biological tools. The work on peptoid nanosheets is particularly exciting as there are very few synthetic struc-tures that self-assemble into planar sheets. To design these self-assembling materials rationally, the hydrophobic effect (a highly important driving force of protein folding) and electro-static interactions were designed to hold together the assem-bled structure. The sequence programmability of the peptoids allowed for the synthesis of structures with various sequence patterns – [(charged)(hydrophobic) x ]y

, but only the alternating sequences self-assembled into nanosheets when mixed. To evaluate which forces are necessary for sheet formation, the charged monomers and then the hydrophobic monomers were replaced with a nonionic hydrophilic monomer. All of these substitutions prohibited sheet formation, con? rming the neces-sity for both the hydrophobic and electrostatic interactions. T he peptoid nanosheets have been characterized with ? uo-rescence microscopy (after the addition of Nile red), SEM, AFM, TEM, and X-ray diffraction. SEM images indicated that most sheets had lateral dimensions on the scale of tens of microns, and revealed sharp, ? at edges of the sheets, sug-gesting that the peptoid oligomers align parallel to those edges. Aberration-corrected TEM was used to visualize the orientation of the peptoids, con? rming their parallel orientation along the edge of the sheet. This again suggested a parallel chain packing geometry consisting of aligned, fully extended chains. An unu-sual mechanism of chain assembly was revealed after detailed studies indicated that a planar air-water interface was required for assembly. The observation that nanosheets only formed in samples that were shaken, not simply stirred, led to the hypoth-esis that these sheets were assembled via organization at the air-water interface into a monolayer, and then compression forced the collapse of the monolayer into a bilayer (Figure 15 b ,d).

[ 171 ] A controlled, reproducible process of vial rocking was devel-oped to allow systematically repeated monolayer adsorption and compression, which allowed for quantitation while avoiding abrasive shaking. The nanosheets formed with this mechanism were larger, but otherwise identical to those previously syn-thesized, as characterized by AFM and X-ray diffraction. The amount of peptoid converted into the nanosheet product was determined to increase linearly with the number of vial rota-tions, supporting the theory that monolayer compression at the air-water interface leads to bilayer formation. This process was additionally characterized via a Langmuir trough, which identi-? ed the pressure required to cause the collapse of the sheets (545.4 mN/m) and the time required to re-establish a surface monolayer of peptoids (450 sec). Interestingly, the sheets were demonstrated to be incredibly stable, and ? uorescence recovery after photobleaching (FRAP) studies revealed that the indi-vidual peptoid chains are relatively immobile over 30 min. C haracterization and analysis of nanosheet formation was continued after determining that a single peptoid chain with both acidic and basic segments (Figure 15 a ) could lead to the

assembly of nanosheets. [ 172 ] To further establish the require-ments for nanosheet formation, self-assembly was evaluated

at the interface between water and hydrophobic solvents.

[ 173 ] In air-free environments, peptoid nanosheets were formed at the interface of water and pentane, hexane, or heptane, but not at the interface of water and more viscous solvents (decane and mineral oil) or aromatic solvents (benzene and toluene). Notably, nanosheets were formed at the water-carbon tetrachlo-ride (CCl 4 )

interface, but they are much smaller. The authors indicate that this is likely due to the high af? nity of CCl 4for aromatic moieties. To investigate the monolayer density being formed at the CCl 4 -

water interface, the surface pressure was

F igure 15. P eptoid nanosheet assembly. a) Structure of a peptoid polymer that leads to nanosheet assembly. b) Schematic of the assembly process due to monolayer formation at the air-water interface followed by compression. c) Schematic of a nanosheet with a loop region that projects from the surface. This loop region can be composed of any peptoid or peptide regions, including the functional region of anti-canine lymphoma IgG2 antibody. A 3D model of the assembled nanosheets is shown d) with and e) without a loop region. f) AFM images of peptoid nanosheets whose loop regions

template the growth of gold nanoparticles after incubation with gold ions. b) Reproduced with permission.

[ 171 ] Copyright 2011, American Chemical Society. d) Reproduced with permission.

[ 98 ] Copyright 2010, Nature Publishing Group. e) and f) Reproduced with permission. [ 176 ]Copyright 2013, American Chemical Society.

REVIEW

measured at both interfaces at various concentrations. The concentration required to saturate the surface monolayer (the maximum surface pressure) was higher for the air-water than the CCl 4

-water interface, suggesting that there is some interac-tion between the peptoids and the carbon tetrachloride which disfavors the collapse into a bilayer. Characterization by X-ray diffraction and AFM con? rmed that all other characteristics of the nanosheets were the same as those formed at the air-water interface. In this work, spectroscopy was used to characterize the sheet formation for the ? rst time. Total internal re? ec-tion vibrational sum frequency spectroscopy was employed to characterize the monolayer formed at the CCl 4 water interface. Side chains were oriented perpendicular to the interface, and ion-pairs between the charged monomers were evident at high concentrations of peptoid. However, if the concentration was decreased and a less dense monolayer was formed, these elec-trostatic interactions were not observed. Similar spectra were observed for a peptoid polymer that lacked the acidic mono-mers. This sequence formed less tightly packed monolayers, likely due to electrostatic repulsion.

A detailed understanding of the nanosheet assembly pro-cess is critical for the future design of functional nanosheets. Analysis of the mechanism of peptoid packing revealed that the

packing of the polymers occurs at the monolayer.

[ 174 ]Grazing-incidence X-ray scattering was used to determine that the monolayer exists as a well-de? ned structure. A coarse-grained molecular model designed to predict the behavior of the back-bone and side chains of the peptoids over a long time-scale was

used to model the monomers at various concentrations.

[ 175 ] The application of this modeling indicated that local ordering of side chain orientation is present in the monolayers at var-ious surface densities of peptoid.

[ 174 ] Increasing the tempera-ture during nanosheet formation leads to an increased rate of assembly and a greater amount of nanosheets formed per vial rotation. This indicates that the formation of the monolayer is the rate-limiting step of nanosheet formation. T hese compiled structural and mechanistic data are critical for the rational design of functional peptoid-based two-dimen-sional materials. Preliminary work has been directed toward the development of af? nity reagents and substrates to template the growth of materials. The ? rst report of functional peptoid nanosheets incorporated a streptavidin binding peptide at the terminus of the peptoid polymer and demonstrated that the assembled nanosheets could bind to a streptavidin-? uorophore

conjugate. [ 98 ] Additional work by Zuckermann et al. has dem-onstrated the incorporation of enzyme substrates and gold-ligands that extend as loops from the plane of the nanosheet

(Figure 15 c ,e). [ 176 ] Peptoid blocks in addition to functional pep-tides were inserted between blocks of alternating hydrophobic and charged monomers. The loops that formed in the mon-olayer were characterized using a Langmuir trough with X-ray re? ectivity capabilities. This indicated that there was additional electron density below the typical peptoid monolayer that was attributed to the loop structure. Nanosheets with protease labile loops were incubated with a commercial protease cocktail, Pro-nase, and then the cleavage was con? rmed by MALDI-TOF MS. Interestingly, however, there appeared to be no disassembly of the nanosheets as characterized by AFM. When nanosheets with the substrate for casein kinase II were incubated with

the protein, phosphorylation was detected with a ? uorescent antibody. As a ? nal demonstration of the ability to incorporate functional loops, nanosheets containing a displayed gold nucle-ation peptide loop were mineralized in the presence of gold ion, and the resulting gold-plated nanosheets were characterized by AFM (Figure 15 f ). The capabilities of these nanosheets high-lights the ability of peptoid-based materials to encode inten-tional protein-like architectural design and functional capacity into the primary sequence of the polymer.

5. M imicking Natural Biological Constructs

5.1. L ipitoids and Surfactants

T he non-viral delivery of nucleic acids into cells is critical for the development of stable, non-toxic, and inexpensive vectors

for nucleic acid based pharmaceuticals. [ 177 ] With the recently

discovered capabilities of the CRISPR/Cas9 system, there seem to be almost limitless applications of these RNA-guided nucle-ases.

[ 178 ] Non-viral nucleic acid delivery is usually achieved by forming a complex with a cationic polymer of lipid. The devel-opment of lipitoids, or lipid-peptoid conjugates, was inspired by peptide-lipid hybrids, which have frequently been employed as

cationic gene delivery agents

[ 179 ] and drug delivery vehicles. [ 180 ] The ? rst lipitoids were polymers of an amine-containing mon-omer attached to various lipids.

[ 181 ] The step-wise synthesis of these structures allowed for a systematic evaluation of how the ratio of charge to hydrophobicity impacts the ef? cacy of the transfection. The peptoids that demonstrated the highest transfection activity also protected DNA from nucleases and contained a repeating trimer of a cationic monomer and two hydrophobic monomers. Interestingly, in a continuation of this work, the transfection ef? ciency could not be directly correlated with size, surface charge, structure (as characterized by CD), or af? nity of the synthetic polymer for DNA (as measured by dif-ferential scanning calorimetry or ITC). [ 182 ] A dditionally, lipitoids have been used as an siRNA transfec-tion agent, ef? ciently transfecting a variety of dif? cult cell lines,

including primary human cells. [ 183 ] The lack of toxicity to immor-talized cell lines was con? rmed and ef? cient siRNA delivery

was demonstrated by monitoring the target genes with Western Blots and real-time PCR. Additional SEM and DLS characteriza-tion was performed in follow-up work to characterize the parti-cles formed from lipitoids and siRNA.

[ 184 ] The size of the parti-cles formed was found to vary depending the ionic strength of the environment and the ratio of lipitoid to siRNA. A nother approach to making functional amphiphilic pep-toids has been extensively investigated by Barron and cow-orkers. Their work has focused on the mimicry of amphiphilic

peptides and proteins using just the peptoid backbone.

[ 185 ] Previous work has mimicked antimicrobial peptides by also varying simple arrangements of cationic and hydrophobic

monomers.

[ 186 ] The Barron group has also pioneered higher molecular weight analogs that mimic the helical and amphi-philic nature of lung surfactant protein C. The ? rst peptoid developed for this purpose was a 22-mer containing a 14-mer of chiral aromatic monomers, two positively charged monomers, and a hydrophobic core. These structures served as a functional

R E V I E W

surfactant without aggregating, demonstrating that a sequence-de? ned peptoid can mimic a 35-mer protein with additional pro-tease resistance. This work was continued by investigating how adding alkyl chains to the N-terminus of this peptoid affected

its activity. [ 187 ] These alkyl chains served as lipid anchors and

increased dynamic ? lm behavior by decreasing the compress-ibility, as measured by pulsating bubble surfactometry. Similar strategies have been applied to creating mimics of the larger

protein, lung surfactant protein B, [ 188,189 ] and the addition

of a lipid to the N-terminus of these molecules had a similar

increased ef? cacy. [ 190 ] Interestingly, peptide-based mimics

of lung surfactant protein B are more effective as dimers, so disul? de and triazole linkages of varying hydrophobicity were

explored to dimerize the peptoid mimetic. [ 191 ]Dimerization

by click chemistry surprisingly led to more surface activity than dimerization by disul? de formation, potentially due to the induction of a hairpin-like turn. The authors hypothesize based on these results that the conformational restraint within the native protein is important for its function. Additionally, the incorporation of achiral hydrophilic monomers into the linker, with 50% aliphatic monomers, led to activity improvement.

5.2. P rotein-Like Polymers and Peptoid-Protein Hybrids

O ne of the ? rst achievements towards mimicking protein structure and function with peptoid polymers was the devel-opment of a protein-like tertiary structure that experienced hydrophobic collapse, leading to the assembly of what was pre-dicted to be a helix bundle.

[ 192 ] In this work, a combinatorial approach was used to identify that monomers on the face of the helix were responsible for driving the assembly. Twelve hydrophobic monomers were introduced in four positions on the 15-mer peptoid to induce the hydrophobic collapse neces-sary for the formation of tertiary structure. The assembly of the identi? ed helices into bundles was followed with circular dichroism (CD). This work was continued by linking together 15-mers through disul? de and oxime linkages to conjugate

the helix bundles covalently. [ 193 ] Fluorophores were attached to

each end of the peptoid polymer to allow for the detection of hydrophobic collapse via FRET quenching. Cooperative disas-sembly was observed upon titrations with various solvents, but the helical structure was unharmed, con? rming that the FRET assay was an indicator of tertiary structure. Understanding the hydrophobic driving forces for the assembly of peptoid architec-tures is important for creating protein-like structures; however, this also fundamentally contributes to the understanding and prediction of protein folding itself. T he fundamental process of hydrophobic chain collapse, the coil-to-globule transition that is common to most protein folding pathways, has also been studied using speci? c pep-toid sequence designs. Work by Zuckermann and coworkers investigated how the sequence of a peptoid polymer containing only one type of hydrophobic monomer and one type of polar monomer contributes to its hydrophobic interactions by syn-thesizing polymers with the same composition but different sequences ( F igure 16a ). [ 194 ] Because computer simulations required a 100-mer chain of de? ned sequence, two 50-mer pep-toids were synthesized and dimerized using click chemistry. One of the structures was designed to be “protein-like” and

contained blocks of hydrophobic ( N -methylglycine) and hydro-philic ( N -(2-carboxyethyl)glycine) monomers designed using

models described by Khokhlov and Khalatur,

[ 195 ] while the other contained repeating blocks of [( N -methylglycine) 4-( N -(2-car-boxyethyl)glycine)]. The stabilities of the assembled globule structures were evaluated by titration with acetonitrile as the unfolding was monitored using SAXS and DLS. The repeating sequence unfolded at lower concentrations of acetonitrile than the protein-like sequence, indicating decreased stability for the repeating structure. This was evidenced by a higher ? uores-cence signal with the protein-like polymer when probed with Nile red (an environmentally-sensitive dye), indicating a more intact globule. The free energy of the folding transition was calculated in addition to the dependence of that energy on the denaturant ( m ), both of which were larger for the protein-like sequence. Notably, a 100-mer peptoid polymer was synthesized in this work – one of the largest sequence-de? ned peptoid polymers that has been synthesized. This also demonstrates the ability to synthesize large polymers using post-synthetic conjugation. It additionally supports the previously mentioned models and demonstrates the utility of sequence controlled polymers in studying fundamental physical properties. A nother way to take on the properties of proteins is to mimic their function. In another report by Zuckermann et al. , a peptoid helical bundle architecture was designed to

mimic a zinc ? nger motif (Figure 16 b ). [ 196 ] Previously identi-? ed helical bundle designs [ 193 ] were modi? ed with zinc-binding

monomer units. The two helices were designed to contain two thirds α-chiral monomers to confer helicity and were built with repeating trimer sequences. The three monomers in the

F igure 16. P rotein-like structures synthesized with peptoid polymers. a)

The hydrophobic collapse of polymers into globular structures in water and acetonitrile mixtures was evaluated for “protein-like” and “repeating” sequences with identical compositions but different monomer arrange-ments. This study indicated that the globular form of the “protein-like” structure was more stable in an organic solvent. b) A peptoid with helix bundle tertiary structure was stabilized with metal coordination. c) A pep-toid was attached to ribonuclease S via an N-terminal serine. a) Repro-duced with permission.

[ 194 ] Copyright 2012, American Chemical Society. c) Reproduced with permission.

[ 198 ] Copyright 2014, ACS.