Geared Toward Applications: A Perspective on Functional Sequence-Controlled Polymers

Abstract

Sequence-controlled polymers are an emerging class of synthetic polymers with a regulated sequence of monomers. In the past decade, tremendous progress has been made in the synthesis of polymers with the sophisticated sequence control approaching the level manifested in biopolymers. In contrast, the exploration of novel functions that can be achieved by controlling synthetic polymer sequences represents an emerging focus in polymer science. This Viewpoint will survey recent advances in the functional applications of sequence-controlled polymers and provide a perspective on the challenges and outlook for pursuing future applications of this fascinating class of macromolecules.

Graphical Abstract

1. Introduction

The ability to synthesize and decode sequence-defined biopolymers such as DNA, RNA, and proteins is the molecular essence of life that enables the evolution of highly sophisticated structures and functions. For example, using transfer RNA (tRNA) as cofactors, the ribosome can decode the genetic information stored in messenger RNA (mRNA) and synthesize polypeptides from amino acids using mRNAs as templates in remarkable efficiency and fidelity. Programmed by their primary sequences, these polypeptides fold into secondary, tertiary, and quaternary structures and can perform a wide array of functions ranging from molecular recognition to catalysis. Aspiring to a myriad of complex functions possessed by biopolymers, generations of polymer chemists took on the formidable challenge to emulate biopolymers and achieve the precise sequence control in synthetic polymers. In his Nobel Lecture, Hermann Staudinger suggested that the studies on synthetic polymers with enhanced structure and sequence control could help to give insight into the structure-function relationship of biopolymers.¹ Several landmark discoveries in polymer science, such as the solid-phase synthesis² and living polymerization,³ were associated with chemists’ efforts to precisely control the sequences and chain structures of macromolecules. In this Viewpoint, we refer to the polymers in which the arrangement of the monomeric building blocks follows a defined order along the polymer chain as “sequence-controlled polymers”.⁴ It is noteworthy that such a definition includes not only monodispersed and polydispersed synthetic polymers with regulated monomer sequences, but also analogs of biopolymers that are monodispersed and consist of extensive non-natural structures and functionalities, such as xenobiotic nucleic acids (XNAs, nucleic acids consisting of non-natural sugar backbones or modified nucleobases^5–6) and peptidomimetics. As a result, the sequence-controlled polymers discussed in this Viewpoint are divided into two main categories: sequence-defined polymers (Ð = 1) and polydispersed sequence-controlled polymers (Ð > 1) (Figure 1). We believe that this broad definition is consistent with the spirit of Staudinger’s macromolecular hypothesis,^7–8 as well as the seminal review by Lutz, Ouchi, Liu, and Sawamoto.⁹ Moreover, it reflects the highly interdisciplinary nature of this field of research and its promising applications in chemistry, biology, and materials science.

Figure 1.

Sequence-controlled polymers discussed in this Viewpoint.

Over the past decade, the development of sequence-controlled polymers was proclaimed as the next “Holy Grail” in polymer science, a topic of paramount importance.¹⁰ New chemical and biological polymerization methods, including iterative coupling, living/controlled chain-growth polymerization, step-growth polymerization, templated polymerization, and enzymatic polymerization, allowed for the preparation of sequence-controlled polymers with diverse sequences and chain structures. Because these advances in the synthetic strategies of sequence-controlled polymers have been comprehensively discussed in many excellent recent research and review articles,^{4, 9–22} this Viewpoint will avoid repeating the discussion on synthesis. Rather, the discussion will be centered on two fundamental questions about function: (1) what novel properties do these sequence-controlled polymers have compared to classic synthetic polymers, and (2) how can these properties address some of the pressing challenges in our society? Hence, this Viewpoint aims to provide a perspective on functional sequence-controlled polymers by highlighting recent examples of the emerging applications of sequence-controlled polymers, such as data storage, biomaterials and biocatalysis, self-assembled nanostructures, and functional soft materials.

2. Functional Applications of Sequence-Controlled Polymers

2.1. Data storage

The amount of data generated and transmitted has experienced an explosive increase since the beginning of the digital age, and it is anticipated that this trend will continue into the foreseeable future. As a result, there is a critical need for new methods and materials that enable efficient and secure data storage. Evidently, DNA as a sequence-defined biopolymer is a promising solution for storing a large amount of information both in living organisms and in laboratories.²³ The defined sequence of the four nucleotides A, G, C, and T gives DNA an information storage capability of 5.5 petabits/mm³, much higher than the traditional silicon-based media. The principle of storing information within the monomer sequence of polymers is not limited to DNA, and sequence-defined polymers should also enable the storage of large amounts of digital information at the molecular level. Compared to DNA, the chemical stability of synthetic polymers offers unique potential advantages for data storage. The ability to incorporate diverse functional groups gives sequence-defined polymers another edge on DNA in applications involving non-biological environments.^{16–17, 24–28} For example, Lutz et al. encoded the 0-bit and 1-bit in binary sequence-defined polyphosphates with degrees of polymerization up to 104.^29–30 By incorporating an alkoxyamine linkage into each phosphoramidite unit, the information-containing polyphosphates were amenable for tandem mass spectrometry (MS/MS) sequencing thanks to the easily cleavable alkoxyamine bonds.³¹ The authors further developed sequence-defined poly(alkoxyamine amides)³² (Figure 2a) and polyurethanes^33–34 (Figure 2b) that contain weak C–ON bonds and C–O carbamate bonds between monomer units, respectively, to allow for higher bond cleavage efficiency by MS/MS. Similarly, Gao and coworkers prepared positively charged sequence-defined polymers using alternating Menschutkin reactions and copper(I)-catalyzed alkyne-azide cycloaddition (CuAAC) reactions in the solution phase.³⁵ The on-source cleavage of the quaternary ammonium group in matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) sequencing allowed for the generation of distinct fragments. Besides the on-source fragmentation within the mass spectrometer, controlled end-to-end chemical degradation has also been used to decode the sequence-defined polymers. For instance, Anslyn and coworkers synthesized sequence-encoded oligourethanes which can be sequenced without the need of MS/MS.³⁶ These oligomers can undergo self-immolation to release 2-oxazolidinones via base-promoted intramolecular cyclization from the terminal alcohol. This process could be monitored by liquid chromatography-mass spectrometry (LC-MS) to provide the sequence information. While these works have demonstrated the potential of sequence-defined polymers in data storage applications, major limitations that require further improvements include: (1) high cost and laborious synthesis limit the amount of information that can be “written” into sequence-defined polymers; (2) high-throughput sequencing of sequence-defined polymers to “read” the stored information remains challenging; (3) unlike DNA, few methods have been developed to date for the amplification of sequence-defined polymers and the information stored in them.

Figure 2.

Sequence-defined polymers for data storage. (a) Oligo(alkoxyamine amide)s consisting of a binary sequence. Reproduced with permission from Ref. 32. Copyright 2015, Springer Nature. (b) The synthesis and MS/MS sequencing of oligourethanes consisting of a binary sequence of 101010. Reproduced with permission from Ref. 34. Copyright 2016, Elsevier Inc.

Data security is another important aspect of data storage. To this end, erasable sequence-defined polymers were developed to undergo structural change in response to external stimuli such as light, temperature, and chemicals. For example, the poly(alkoxyamine amides)s reported by Lutz et al. are temperature-labile and able to degrade at elevated temperature (120 °C) in the presence of a large excess of 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), resulting in the erasing of the encoded information.³² They also decorated the aforementioned phosphoramidite monomers with light-labile o-nitrobenzyl/o-nitroveratryl ether and light-inert p-nitrobenzyl ether side chains and assembled them into photo-editable binary encoded polymers.³⁷ These polymers can be read out using MS/MS but became unreadable after UV exposure. Synthetic sequence-defined polymers also offer the possibility to be used for anticounterfeiting and cryptography. For example, a small amount of binary encoded oligourethanes synthesized by Lutz et al. could be blended into industrial plastics such as polystyrene and poly(methacrylate).^33–34 The information stored in these oligomers can be extracted from the mixed matrix by selective dissolution in methanol or tetrahydrofuran followed by MS/MS decoding. To enhance the information-carrying capacity, the authors further developed a 2D coding approach by writing information in a set of binary coded oligo(alkoxyamine amide)s. The selective identification by LC-MS and MS/MS could be used to decode the information stored in these 2D barcodes.³⁸

2.2. Bioactivities

2.2.1. Recognition of biomolecules

Glycan recognition by receptors plays important roles in a variety of biological activities such as cell-cell interaction, signal transduction, and host–pathogen recognition.³⁹ To elucidate the relationship between the structure of glycan ligands and their receptor-binding properties, glycomimetic sequence-defined polymers have been developed as synthetic analogs of natural polysaccharides. For example, Hartmann et al. developed the sequence-defined glycopolymers carrying different sugar ligands including mannose (Man), galactose (Gal) and glucose (Glc) at precise positions on the peptide backbone (Figure 3a).^40–41 They found variations in both the number and sequence of the sugar ligands could alter the binding affinity. Specifically, the presence of non-binding galactose and glucose moieties could suppress receptor clustering and promote steric shielding of the heterotrivalent glycopolymer, and thus enhanced their affinity to Concanavalin A (Con A). Polymer-receptor binding has also been studied in the context of polydispersed sequence-controlled glycopolymers. For instance, Haddleton et al. prepared a library of multiblock glycopolymers with block sequences consisting of mannose, glucose, fucose side chains via atom transfer radical polymerization (ATRP) (Figure 3b).⁴² The dissociation constants (K_D) and half maximal inhibitory concentration (IC₅₀) of these glycopolymers to dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN) exhibited a general correlation to the mannose content. Collectively, these examples demonstrated that sequence control in both sequence-defined polymers and polydispersed sequence-controlled polymers can significantly influence the complex, multivalent nature of the interactions with proteins.

Figure 3.

Biomolecular recognition by sequence-controlled polymers. (a) Glycopolymers with heteromutivalent and homomultivalent structures consisting of mannose (Man), glucose (Glc) and galactose (Gal) side chains at defined positions for lectin binding. Reproduced with permission from Ref. 40. Copyright 2014, American Chemical Society. (b) Multiblock glycopolymer capable of inhibiting the DC-SIGN-gp120 recognition. Reproduced with permission from Ref. 42. Copyright 2013, WILEY-VCH. (c) SOMAmers bind their targets with exquisite shape complementarity and utilize hydrophobic modifications at the binding interface. Reproduced with permission from Ref. 47. Copyright 2014, Elsevier Inc. (d) Click-PD strategy for two-color screening of lectin-specific base-modified aptamers and FACS (fluorescence-activated cell sorting) plot with two-color screening for affinity and specificity in parallel. Reproduced with permission from Ref. 51. Copyright 2019, American Chemical Society.

XNA-based affinity reagents, often referred to as “XNA aptamers”, are a class of sequence-defined polymers extensively studied for binding to biomolecular targets. Two important advantages of XNA aptamers compared to their DNA/RNA counterparts are the ability to incorporate customized functional groups to optimize binding affinity according to its applications and the resistance to nuclease-mediated degradation in vivo. Romesberg et al. identified XNA aptamers with 2’-fluoro and 2’-OMe substitutions of the sugar backbone binding to human neutrophil elastase (HNE), a serine protease associated with inflammatory diseases.⁴³ Building upon their work on polymerase engineering, Holliger et al. prepared a 2′-deoxy-2′-fluoroarabino nucleic acid (FANA) library and selected FANA aptamers targeting HIV-1 reverse transcriptase (RT) and integrase (IN), respectively, via systematic evolution of ligands by exponential enrichment (SELEX), with affinities comparable to DNA and RNA aptamers.^44–45 While modifications on the sugar backbone could improve XNA’s resistances to nucleases, side chain modifications on the nucleobases were explored to expand the scope of functionalities in aptamers. Eaton et al. developed a 5-modified dUTP derivatives that incorporate a series of hydrophobic groups through an amide linkage and incorporate these 5-modified dUTP derivatives in an XNA library for in vitro selection.⁴⁶ Comparably, the incorporation of hydrophobic/aromatic functional side chains at the 5-position of uracil generated slow dissociating aptamers to a wider range of protein targets and enhanced the success rate of SELEX. These so-called SOMAmers (Slow Off-rate Modified Aptamers) utilize their hydrophobic modifications at the binding interface to complement the target protein surfaces that are significantly more hydrophobic, increasing their binding potential to a broader scope of target proteins (Figure 3c).⁴⁷ Using the T4 DNA ligase-mediated DNA-templated polymerization strategy, Liu et al. and Hili et al. independently generated highly functionalized nucleic acid polymer (HFNAP) libraries and identified HFNAP aptamers targeting therapeutically relevant proteins such as PCSK9, interleukin-6, and thrombin via SELEX. These HFNAP aptamers feature low nanomolar affinity and high selectivity, highlighting the importance of expanding the chemical space of side chain modifications in XNA aptamers.^48–50 Soh, Niu, and coworkers combined a CuAAC click chemistry-based DNA modification strategy with particle display (PD) screening platform to select carbohydrate-modified aptamers for the recognition of carbohydrate-binding proteins (Figure 3d).⁵¹ Specifically, emulsion polymerase chain reaction (PCR) was performed to produce monoclonal aptamer particles in which the thymine base was replaced by an alkyne-modified uridine. Azide-modified mannose was subsequently conjugated to the aptamer particles via CuAAC. The carbohydrate-modified aptamer particles were screened for binding to fluorescently labeled lectin targets in a two-color fluorescence-activated cell sorting (FACS). The selected aptamers exhibited low nanomolar affinity and exceptional specificity to ConA, with minimal binding to other structurally similar lectins. In another notable example, Krauss et al. developed DNA-supported clusters of carbohydrates through selection with modified aptamers (SELMA) to mimic the epitope of 2G12, an antibody that protects against HIV infection.^52–55 This technology involves using DNA as cluster scaffolding materials and sequence-specifically introducing glycan modifications such as dense high mannose glycans (Man₄) via CuAAC coupling. Glycoclusters discovered in SELMA displayed low nanomolar affinity to 2G12. These glycoclusters have shown promising implications for the design of future drugs and vaccines.

Sequence-defined XNA oligomers have also been applied to binding to DNA/RNA in vivo and inhibiting gene expression. Known as antisense oligonucleotides, they function through forming sequence-specific complexes with the target DNA or RNA and disrupting their replication or transcription. A variety of XNA backbones have been studied for antisense applications, including locked nucleic acid, peptide nucleic acid, morpholino nucleic acid, phosphorothioate nucleic acid, etc.^56–58 Compared to DNA- and RNA-based reagents, XNA-based antisense oligonucleotides possess improved stability, pharmacokinetics, and antisense activity in vivo. To date, several XNA-based antisense oligonucleotide drugs, such as Vitravene (fomivirsen) and Kynamro (mipomersen sodium), Inotersen (Tegsedi), Eteplirsen (Exondys 51), golodirsen (Vyondys 53) have already received FDA approval, with several others undergoing clinical trials.^59–60

Peptidomimetics such as peptoids, β-peptides, and γ-peptides are another class of sequence-defined biomimetic polymers used for biomolecular recognition. Compared to peptides, advantages of peptidomimetics include high proteolytic stability, low mammalian cytotoxicity, and the high flexibility of incorporating functional groups. Zuckermann and coworkers first synthesized peptoid libraries and screened them for their binding abilities to protein targets.⁶¹ Since then, a plethora of peptoid ligands to therapeutically relevant proteins has been discovered. For example, Kodadek et al. identified a peptoid oligomer that can bind to biomarkers related to Parkinson’s disease,⁶² HIV,⁶³ Alzheimer’s disease,⁶⁴ and other diseases.⁶⁵ In addition, Seo et al. developed functional glycopeptoids through N-alkylaminooxy glycosylation. Serving as glycopeptide mimics or carbohydrate-presenting materials, these glycopeptoids created opportunities for revealing new insights into carbohydrate-protein interactions.⁶⁶ Schepartz et al. systematically studied β-peptides as protein inhibitors. They reported a β³-peptide 14-helix structure that can inhibit the human oncogene product double minute 2 (hDM2), an important cancer therapy target, with nanomolar affinity.⁶⁷ Later on, they developed β-peptides that contain diether and hydrocarbon bridges to improve cell permeability and circumvent the need for an Arg₈ sequence for cellular delivery.⁶⁸ Besides inhibiting the p53-hDM2 interactions, β³-peptide has also been developed to serve as HIV fusion inhibitors. They further developed a β³-decapeptide with a non-natural trifluoromethylbenzene side chain that inhibited the HIV gp41-mediated fusion with a CC₅₀/EC₅₀ ratio of 8, a major improvement compared to previous works.^69–70 Cai et al. developed an impressive class of sulfono- γ -AApeptides that can reproducibly fold into well-defined helical structures.^71–73 They demonstrated that these sulfono- γ -AApeptide helices were excellent mimicries of α-helices in several therapeutically relevant protein-protein interactions and could serve as inhibitors or agonists to these interactions, such as BCL9-β-catenin, p53-MDM2/MDMX, and glucagon peptide-GLP-1R.^74–77

2.2.2. Antimicrobial and antifouling applications

Antimicrobial peptides (AMPs) have attracted significant research and development activities given their wide-spectrum efficacy against bacterial pathogens and the difficulty for pathogens to develop resistance. However, limitations of AMPs such as poor scalability, high cost, and protease degradability have impeded their applications. To address these issues, a wide range of synthetic mimics of AMPs composed of cationic sequence-controlled polymers have been developed.^78–82 These antimicrobial polymers often carry cationic and hydrophobic functional groups. Their antimicrobial activities arise from the electrostatic interactions between the cationic side chains and the negatively charged phospholipids in the bacterial cell membranes, and the membrane disruption by the hydrophobic functionalities of the bound polymers.^83–84 As a result, the antimicrobial activities of these polymers are strongly influenced by the sequence, charge, hydrophilic/hydrophobic moiety ratio, molecular weight, and self-assembled morphologies. Baran and coworkers prepared 27 peptoids with antimicrobial structures and found a number of these peptoids had potent antimicrobial activities against a wide range of Gram-positive and Gram-negative strains.⁸⁵ Beyond peptoids, other synthetic sequence-defined polymers have also been discovered to possess antimicrobial activities. For instance, Alabi and coworkers synthesized a unique class of sequence-defined oligothiotheramides from allyl acrylamide monomers using liquid-phase iterative synthesis and fluorous tag-facilitated purification.^78–80 The precise sequence control in these oligothiotheramides enabled the precise tuning of the hydrophobicity and the antimicrobial activity. Based on the mechanism of action of the antimicrobial polymers discussed above, having a defined sequence and monodispersity may not be a prerequisite for antimicrobial activities. Indeed, Boyer et al. have shown that polydispersed multiblock copolymers could also be mimics of antimicrobial peptides. They prepared a library of 32 multiblock copolymers consisting of cationic, hydrophilic, and hydrophobic blocks and screened the library against a wide range of bacterial pathogens.⁸¹ The results confirmed that by tuning the distribution of monomers within blocks a balance between the desired antimicrobial activity and undesired hemolytic activity could be achieved.

Biofouling is a process in which living organisms and organic matter produced by the deposit on surfaces.⁸⁶ Biofouling is highly undesirable in many cases such as adhesion of marine organisms on the ship hull and biofilm formation on the surface of biomedical devices. Antifouling sequence-controlled polymers have been developed as coating materials to render the modified surfaces resistant to fouling by proteins and bacteria. Segalman et al. conjugated a series of sequence-defined polypeptoids composed of two monomers, a hydrophilic N-(2-methoxyethyl)glycine and a hydrophobic N-(heptafluorobutyl)glycine, to the surface to prevent the biofouling by the algae Ulva linza.⁸⁷ They found that the position of fluorinated monomer in the polypeptoid sequence imposed a major influence on the antifouling property of the coating against Ulva linza. Messersmith et al. prepared a library of zwitterionic peptoid brushes to evaluate their antifouling properties.⁸⁸ Their results showed that at a sufficiently low ionic strength, the polarity of the charged residues near the termini of the polypeptoid chains determined the electrostatic adsorption of the proteins to the surface.

2.2.3. Drug delivery

Biocompatible polymers are well known as essential carriers to deliver drugs or genes for therapeutical applications. One of the advantages of using polymeric carriers is that the intracellular trafficking of molecular cargos could be regulated by macromolecular architectures including composition, topology, and self-assembled morphologies. Tew et al. prepared three synthetic polymers to mimic the protein transduction domains (PTDs) via ring-opening metathesis polymerization (ROMP).^89–90 These homopolymers, gradient copolymers, and block copolymers had the same length and overall composition of cationic (guanidine) and hydrophobic (phenyl) groups, but had different degrees of segregation of hydrophobic side chains. While all of them could be internalized in Jurkat-T cells and HEK293T cells, the gradient copolymer with an intermediate degree of functional group segregation had the optimal balance of internalization activity, solubility, and toxicity. Perrier et al. investigated the influence of monomer sequence on the cellular uptake of polymers, which are multiblock polymers consisting of guanidinium-functionalized acrylamide and two other less hydrophilic monomers. These studies demonstrate that the sequence of blocks has an influence on both the uptake efficiency and the mechanism of internalization.⁹¹ Wagner et al. synthesized a library of oligo(ethane amino) amides with precise modification patterns and topology to study the structure-activity relationship between the polymer sequence and nucleic acid delivery activity.^92–97 This work indicated that the optimization of the location of hydrophobic and crosslinkable functional groups can improve the siRNA polyplex stability and transfection efficiency. Two candidates of a 38-oligomer library have promising properties of siRNA-binding and gene-silencing activities and achieved excellent efficacy in inhibiting tumor growth in vivo when formulated with therapeutic EG5 siRNA. In another example, Zuckermann et al. developed “lipitoids” composed of a cationic peptoid oligomer and a hydrophobic lipid moiety for siRNA delivery.^98–99 These lipitoids can self-assemble into spherical nanoparticles with siRNA. It was found a particular repeating trimer sequence of cationic-hydrophobic-hydrophobic demonstrated high transfection efficiency and low cytotoxicity.¹⁰⁰ However, no correlation was found between the transfection efficiency and the structures of the DNA-peptoids/lipitoids complexes, suggesting that further investigation is needed to probe the hidden structure-function relationships.¹⁰¹

2.2.4. Catalysis

Similar to DNA and RNA that can form ribozymes and DNAzymes, single-stranded XNAs can be selected to fold into particular motifs and serve as molecular catalysts. Furthermore, modifications to the sugar backbone and nucleobases provided the XNA catalysts (XNAzymes) superior stability and/or enhanced catalytic activities. Holliger et al. presented an impressive example of XNAzyme with RNA endonuclease and ligase activities.¹⁰² They subjected chimeric DNA–XNA libraries of four types of XNA (arabinonucleic acid (ANA), FANA, hexitol nucleic acid (HNA), cyclohexyl nucleic acid (CeNA)), all derived from DNAzyme motifs, into in vitro selections for RNA cleavage and ligation. While the activities varied, the selected XNAzymes all exhibited RNA cleavage/ligation activities. They also used the same strategy to identify the first successful XNA-cleaving FANAzyme with a rate constant of 0.02 min⁻¹. Chaput et al. developed an RNA-cleaving FANAzyme with higher catalytic efficiency (k_cat = 0.2 ± 0.01 min⁻¹) and Michaelis-Menten kinetics (Figure 4a).¹⁰³ The discoveries of these XNAzymes proved that catalytic activities can emerge from synthetic XNA polymers, a general principle that can also be potentially expanded to other types of sequence-defined synthetic polymers.

Figure 4.

Catalytic XNAs and peptidomimetics. (a) Cleavage of a chimeric DNA/RNA substrate contains a single unpaired ribo-G residue by FANAzyme NGS12–7. Reproduced with permission from Ref. 103. Copyright 2018, Springer Nature. (b) The foldamer catalyst for macrocycle formation. Reproduced with permission from Ref. 105. Copyright 2019, AAAS.

Molecular catalysts based on peptidomimetics have also been developed. Kirshenbaum et al. synthesized a series of TEMPO conjugated peptoid helices to achieve the catalytic function of enantioselective alcohol-ketone transformation.¹⁰⁴ They discovered that the enantioselectivity of the catalytic peptoids depends on the handedness of the helical scaffold, the position of the catalytic center along the peptoid backbone, and the folding of the peptoid scaffold. In another example, Gellman et al. reported a peptidomimetic foldamer containing both α- and β-amino acid residues to be used as catalysts on the ring-closing in macrocycles synthesis. When folding into a three-dimensional helix structure, this foldamer presents a primary amine and a secondary amine in defined positions to promote aldol reactions that form rings containing 14 to 22 atoms (Figure 4b).¹⁰⁵

2.2.5. Protein mimics

Empowering synthetic polymers with protein-like structures and functions has been a longstanding goal for polymer chemists. Recent advances in peptidomimetic polymers have shed light on this important research field. Zuckermann et al. designed a peptoid helix bundle structure equipped with thiol and imidazole handles that can bind to zinc cations and bridge two helix bundles in a tertiary structure. Such a structure resembles the zinc-binding motif of zinc-finger domains found in DNA-binding proteins.¹⁰⁶ Schepartz et al. studied the development of β-peptide bundles that mimic protein functions including carbohydrate recognition, metal binding, and catalytic activity.¹⁰⁷ For example, they incorporated β-amino acid analogs with a boronic acid side chain into the β-peptide bundle to achieve high affinity to sorbitol, highlighting the potential to develop artificial β-peptide lectins in the future.¹⁰⁸ Similarly, incorporating β³ -homocysteine (β³ C) residues in β-peptide bundles enabled selective Cd²⁺ ion binding activity over other metal ions tested (Hg²⁺, Pb²⁺, and Zn²⁺).¹⁰⁹ They also designed a β-peptide bundle catalyst aiming to operate as esterases. The β-peptide bundle sequence was optimized to increase the affinity to the substrate, resulting in catalytic efficiency of k_cat/K_M = 98 M⁻¹ min⁻¹. Structural studies of the β-peptide bundle esterase reveal that bundle formation plays a critical role in catalysis, paving the road for future research on polymer-based protein mimics.¹¹⁰

2.3. Self-Assembled Nanostructures

Through judicious design, sequence-controlled polymers can self-assemble into nanostructures that have promising prospective applications in fields of lithography, drug delivery, sensors, and optoelectronics. Studies on these sequence-controlled polymers will also reveal important insights into the relationship between monomer sequences and the morphologies of polymer self-assembly.

2.3.1. Self-assembly of sequence-controlled synthetic copolymers

Johnson et al. prepared monodisperse diblock 32-mer copoly(triazoles) with different stereochemical sequences via iterative exponential growth (IEG). The group found that different monomer stereochemistry and sequences led to dramatic changes in polymer self-assembly (Figure 5a).¹¹¹ Small-angle X-ray scattering (SAXS) and transmission electron microscopy (TEM) analyses showed that L/L diblock 32-mer, D/L diblock 32-mer, and alt/L diblock 32-mer yielded three distinct morphologies: double-gyroid, lamellar, and hexagonal cylinder, respectively. The incorporation of a small-molecule additive (L)-tartaric acid can swell the hydrophilic block and further induce the morphology transformation from gyroid to lamellar for L/L 32mer, and hexagonal cylinder to lamellar for alt/L 32mer. In another example of the self-assembly of sequence-defined polymers, Cheng et al. prepared a small library of polymers composed of two types of giant molecules polyhedral oligomeric silsesquioxane (POSS): hydrophilic DPOSS and hydrophobic BPOSS.¹¹² The collective hydrogen bonding and hydrophobic interactions induced the nanophase separation. Tailoring the location of DPOSS and BPOSS in the polymer backbones results in different well-defined supramolecular lattices including lamellar, double gyroid, hexagonal cylinder, Frank–Kasper (F-K) A15, and body-centered cubic. Regulating sequence in polydispersed sequence-controlled polymers has also been used to tune their assembled structure in the bulk interface of two phases. For instance, Coates et al.^113–114 utilized polyethylene/isotactic polypropylene (PE/iPP) multiblock copolymers with precise control over block length as the compatibilizer to weld common grades of commercial PE and iPP together. With comparably short block length of PE and iPP, tetrablock copolymers PP₃₆PE₂₀PP₃₄PE₂₄ exhibited better adhesive strength than diblock PP₂₄PE₃₁ due to the formation of entangled loops in the interfaces. Simmons et al. employed a molecular-dynamics simulation-based genetic algorithm to predict the performance of sequence-specific copolymers as a compatibilizer. They showed that sequence-specific copolymers offered the potential to yield a larger reduction in interfacial energy than either block or random copolymers.¹¹⁵ In another example, Cai et al. investigated the “sequence-controlled polymerization-induced self-assembly” of block polyions in aqueous media.¹¹⁶ Three types of block polyions with different sequences of ionic monomers, i.e., block, gradient, and alternating zwitterionic sequences, were prepared via photoswitchable reversible addition-fragmentation chain-transfer (RAFT) copolymerization using polyethylene glycol chain transfer agents. While block zwitterionic copolymerization resulted in precipitation, and polyions with alternating zwitterionic copolymerization were water-soluble, various gradient zwitterionic sequences led to the formation of different nanostructures such as spheres, lamellae, and vesicles. These studies confirmed that polymer sequences can significantly affect inter- and intramolecular interactions, and effectively mediate the self-assembled nanostructures in bulk and in solution.

Figure 5.

Self-assembled nanostructures formed by sequence-controlled polymers. (a) Unimolecular stereoisomeric diblock 32-mers synthesized by IEG. Their 1D intensity SAXS and WAXS plots indicate distinct phases. Reproduced with permission from Ref. 111. Copyright 2018, American Chemical Society. (b) 3D structure of tetrahedral assembled from AuNP-labelled XNA strands confirmed by TEM at different tilting angles. Reproduced with permission from Ref. 117. Copyright 2016, Wiley-VCH. (c) Antibody-mimetic peptoid nanosheets for molecular recognition. Reproduced with permission from Ref. 120. Copyright 2013, American Chemical Society.

2.3.2. XNA nanostructures

It is well established that natural nucleic acids such as DNA are highly programmable and can fold into complex 3D structures. Such programmability and complex nanostructure assembly have recently been extended to XNAs. For example, Holliger and coworkers described XNA-based 70 kDa Turberfield tetrahedron nanostructures from four different classes of XNAs: 2’F-RNA, FANA, HNA, and CeNA (Figure 5b).¹¹⁷ Impressively, despite their conformational heterogeneity, all XNAs successfully formed the same tetrahedra. A 600 kDa all-FANA octahedron was also assembled. Further analysis of the FANA octahedron structure revealed differences in shape and structural orientation compared to that of natural DNA, which may be due to the conformational differences in sugar puckering between FANA and DNA. Notably, the full XNA nanostructures were found to be significantly more stable than those made of natural DNA in the presence of nucleases.

2.3.3. Peptoid nanostructures

The facile synthesis and functionalization make peptoids a great material for building self-assembled nanostructures. Zuckermann et al. discovered a class of polypeptoids that can fold into water-soluble and precisely oriented nanosheets.¹¹⁸ The self-assembly of these polypeptoids formed a highly ordered, planar bilayer structure that could extend to hundreds of units long. This process was driven by the phase separation of the hydrophobic and charged monomers in the same peptoid sequence.¹¹⁹ These peptoid nanosheets had a chemically defined interior and exterior that were only 3 nm thick, and could be custom modified with various functional groups on the exterior surface. The authors further demonstrated the utility of these nanosheets by fabricating an antibody-mimetic two-dimensional substrate. Based on the structural component of natural antibodies, they decorated the exterior surface of the nanosheets with antibody-mimicking peptide loops. The high-surface area of the nanosheets served as a scaffold for a variety of functional peptide loops that could be used for binding to target biomolecules (Figure 5c).^120–121 Cryogenic transmission electron microscopy imaging of the crystal lattice of the diblock copolypeptoid nanosheets revealed further insights into how the aromatic side chains in the polypeptoid sequence influenced the crystal structure of the nanosheet crystals at the atomic level.¹²² Similarly, Zhang et al. found that progressively placing positive charge away from the hydrophobic block led to the increase of the micellar radius and aggregation number in the micelles assembled by amphiphilic ionic block copolypeptoids.¹²³ Zuckermann et al. demonstrated that the microphase separation behavior of sequence-defined peptoid diblock copolymers in the solid state differed from the traditional block copolymers.¹²⁴ In particular, only lamellar and disordered morphologies were observed over the entire window of the composition and temperature variation. Furthermore, the maximum order-disorder transition temperature was reached when the volume fraction of the hydrophilic block was 0.24 in these peptoid diblock copolymers, compared to 0.5 for the same parameter in conventional block copolymers.

2.4. Macroscopic Material Properties

The ability to regulate the sequence along the polymer chain created an important opportunity to fine tune the macroscopic properties of sequence-controlled polymers including degradability,¹²⁵ thermal properties,^126–128 mechanical properties,¹²⁹ optoelectronic properties,^130–133 etc. Meyer et al. investigated the impact of sequence on the degradation of poly(lactic-co-glycolic acid) (PLGA).¹²⁵ They observed that, under neutral phosphate buffer, random copolymers showed exponential decay of 70% weight loss in four weeks, while the weight loss of the alternating copolymer was considerably smaller over the same period. After a rapid initial weight loss, the degradation of the alternating copolymers exhibited a near zero-order behavior. The authors attributed the slower degradation rates of the alternating copolymers to their unique sequence: the alternating copolymer has only one connectivity pattern, i.e., lactic-glycolic, while the random copolymers have two more connectivity patterns, i.e., lactic-lactic and glycolic-glycolic (Figure 6a). Since it was expected that the degradation rate followed the order of glycolic-glycolic > lactic-glycolic > lactic-lactic, the random copolymer degraded faster at the beginning and gradually slowed down. In contrast, the degradation rate of alternating copolymer remained steady over the course of the reaction.

Figure 6.

Macroscopic material properties of sequence-controlled polymers. (a) Sequence-controlled PLGA copolymers demonstrated distinct hydrolytic degradation rates compared to random copolymers; Reproduced with permission from Ref. 125. Copyright 2011, American Chemical Society. (b) The sequence of polyurethane oligomers has an influence on network topology and mechanical properties upon crosslinking using dithiol linkers. Reproduced with permission from Ref. 129. Copyright 2020, American Chemical Society. (c) Diastereo- and enantioselective ROP of a diastereomeric mixture of rac/meso-8DL^Me could generate diblock stereocopolymers it-sb-st-P3HB consisting of it- and st-stereoblocks of P3HB. Reproduced with permission from Ref. 136. Copyright 2019, AAAS.

Alabi et al. reported a clever strategy for the scalable production of sequence-defined polyurethane networks (Figure 6b).¹²⁹ By utilizing highly efficient reductive amination and amine-anhydride condensation reactions, large scale synthesis of sequence-defined polyurethane macromers (SD-PUM) was readily achieved. Specifically, SD-PUM with three different sequences AAmmmA, mAmAmA, and AAAmmm (where A and m represent the units with allyl group and methyl group on urethane, respectively) were prepared and crosslinked by dithiol crosslinkers. It was revealed that polyurethane networks with AAAmmm sequence always showed lower T_g than mAmAmA- and AAmmmA networks. They also exhibited ~40% and ~30% lower rubbery modulus than that of mAmAmA- and AAmmmA networks. The authors explained that the relative amounts of intramolecular loops and dangling chain ends may account for different thermal and mechanical properties.

The stereosequence also plays an important role in influencing polymer properties. A typical example is that atactic poly(propylene succinate) synthesized by alternating copolymerization of racemic propylene epoxide and succinic anhydride showed T_g of −39 °C,¹²⁶ while the isotactic poly(propylene succinate) from enantiopure propylene epoxide had a much lower T_g of −4 °C. This shows the huge influence of stereosequence on polyester thermal properties.¹³⁴ Recently, Chen et al. reported an ingenious method to make stereosequenced polyhydroxyalkanoates from diastereomeric monomer mixtures.¹³⁵ By using highly stereoselective catalysts, rac (R,R and S,S) or meso (R,S) cyclic diolide (8DL^Me, Me denotes the methyl group on the 8DL ring) were polymerized both enantioselectively and diastereoselectively to form isotactic and syndiotactic poly(3-hydroxybutyric acid) (P3HB), respectively (Figure 6c).¹³⁶ In addition, the polymerization of rac-8DL^Me was faster than that of meso-8DL^Me, thus the synthesis of a tapered stereodiblock polymer it-P3HB-sb-st-P3HB was possible. By changing the rac/meso ratio, it-P3HB-sb-st-P3HB with similar molecular weights (16.8–17.3 kg/mol) showed different T_m values. Changing the methyl group on monomer into an ethyl group allowed for the synthesis of more complex tapered copolymers with improved mechanical properties. These results clearly demonstrated the significance of stereosequence in determining the polymer properties.

3. Conclusion and Outlook

Since Staudinger’s seminal paper Über Polymerisation⁷ a century ago, polymer chemists have made great strides in controlling the molecular weight, dispersity, sequence, microstructure, and chain-end groups of synthetic polymers. Evidenced by this Viewpoint, the study on the sequence-controlled polymers is a burgeoning and highly interdisciplinary field featuring the convergence of many previously independent research topics such as polymer chemistry, biochemistry, glycoscience, nucleic acid chemistry, protein engineering, immunology, and information technology. The integration of ideas and expertise from these diverse disciplines helped to create sequence-controlled polymers that not only exhibited improved functionalities in the existing applications of synthetic polymers, but also stimulated novel applications inaccessible by traditional polymers, such as data storage, molecular recognition, programmable self-assembly, vaccine, and catalysis. Particular applications often require specific molecular features and degree of sequence control. For example, the application of data storage requires high polymer molecular weight, defined building block sequence, and monodispersity to ensure high data storage capacity and fidelity. Biological applications such as molecular recognition, antimicrobial activity, and drug delivery need the sequence-controlled polymers to be water soluble or amphiphilic. Cationic functionalities are often incorporated in the sequence-controlled polymers used for the antimicrobial application and RNA delivery due to the need for the electrostatic interaction with negatively charged molecules. As advancements continue to emerge, there are great opportunities for the development of new sequence-controlled polymers to provide improved or completely new functions in the future.

New functionalities of sequence-controlled polymers are synergistically connected with the development of new synthetic approaches. For instance, methods that can significantly increase the molecular weight and the scale of synthesis of sequence-defined polymers are essential for their information-carrying ability to approach that of DNA in data storage applications. Toward these goals, the IEG strategy developed by Johnson et al.^{111, 137} and the single unit monomer insertion (SUMI) approach developed by Xu and Boyer^138–140 have demonstrated promising potentials (Figure 7a), but further efforts are still needed: while the density of variable monomer units incorporated by IEG requires improvement, SUMI is still limited to short oligomers. Furthermore, compared to DNA, another limitation of existing sequence-defined polymers is their inability to undergo a replication and amplification process analogous to PCR of DNA, making it challenging for the decoding, amplification, and transmission process of the information carried in these polymers. New methods that can allow for replication and amplification of synthetic polymers, exemplified by recent advances in XNA technologies⁶ and DNA-templated polymerization,¹⁴¹ will offer important advantages in data storage over existing sequence-defined polymers.

Figure 7.

Strategies to access sequence-controlled polymers with more complex structure and functions. (a) Schematic illustration of the single unit monomer insertion (SUMI) approach to access sequence-defined oligomers. Reproduced with permission from Ref. 139. Copyright 2017 Wiley-VCH. (b) Covalent folding of linear synthetic polymer chains could be achieved through strategic positioning of crosslinkable side chains in the polymer structure. Reproduced with permission from Ref. 142. Copyright 2011, Springer Nature. (c) Three methods for the sequence identification of functional sequence-controlled polymers: bioabstraction, combinatorial screening, and rational design. Reproduced with permission from Ref. 13. Copyright 2019, Royal Society of Chemistry.

One of the ultimate goals of the studies on sequence-controlled polymers is to generate polymers that can fold into protein-like globular structures and execute sophisticated chemical and biochemical functions such as receptor recognition and catalysis. Toward this goal, two distinct, but converging strategies have been adopted: (1) controlling the folding and the positioning of functional groups of single-chain polymers by regulating their sequence and composition, and (2) engineering existing biopolymers or recapitulating the structures of biopolymers by their synthetic mimics. An excellent example of the former approach was reported by Lutz et al., in which they added discrete amounts of functionalized maleimides at various time points during the ATRP polymerization of styrene to precisely introduce crosslinkable functional groups through the incorporated maleimides in the polymer chain, allowing for controlled polymer folding (Figure 7b).¹⁴² The latter approach has been explored by many researchers, yielding promising results thus far. The XNA nanostructures developed by Holliger¹⁴³ and Chaput,⁵⁹ and the peptoid and β-peptide bundles reported by Zuckermann¹⁰⁶ and Schepartz,¹⁰⁷ respectively, are notable examples. One particularly noteworthy direction that holds a great promise to generate protein-like synthetic polymers is to engineer the ribosome as a polymer synthesis machine. This formidable task may require many iterations of mutagenesis and structural optimization to the natural ribosome-tRNA machinery, but the seminal work by Fahnestock and Rich,¹⁴⁴ and recent reports by Suga,^145–147 Schepartz,^148–150 and Jewett¹⁵¹ have shown important advances in engineering the ribosome-tRNA machinery to incorporate non-natural β - and γ -amino acids in polypeptides as well as α -hydroxyacids in polyesters.

In the process to generate functional sequence-controlled polymers, knowing what sequence to make is equally if not more important than developing the methods to make it.  Three distinct strategies: “bioabstraction”, rational de novo design, and selection/screening from combinatory libraries were originally summarized by Brummelhuis, Wilke, and Börner for identifying functional peptides,¹⁵² and were further elaborated by Austin and Rosales in a recent review for identifying sequence-controlled polymers (Figure 7c).¹³ “Bioabstraction” implies the direct translation of the biopolymer sequence into the sequence of synthetic polymers, which has been widely adopted to generate biomimetic polymers such as the peptidomimetic affinity reagents discussed in Section 2.2.1 and the antimicrobial polymers discussed in Section 2.2.2. Capitalizing on the recent advances in de novo protein design¹⁵³ and artificial intelligence,¹⁵⁴ the de novo design of sequence-controlled polymers has few published examples to date but has emerged as a highly active area of research widely anticipated to evolve rapidly in the near future. A significant obstacle to this strategy is that unlike the substantial databases of proteins, existing knowledge on the sequence-structure-functional relationship of synthetic polymers is few and far between, making it challenging to develop training sets for the computational algorithms. Therefore, further developments of the de novo design strategy for discovering functional sequence-controlled polymers will likely require inputs from new experimental approaches, as well as adapt from existing computational platforms for biopolymer structural prediction.¹⁵⁵ Perhaps one of the most promising experimental approaches to discover functional sequence-controlled synthetic polymers is the selection/screening from combinatorial libraries. Reported approaches in this direction range from the screening of one-bead-one-compound (OBOC) library¹⁵⁶ to the selection of DNA-encoded polymers.¹⁴¹ In all of these methods, with the exception of some XNA and peptidomimetics that can be sequenced directly, a key step is to tag each member of the combinatory library with a unique molecular barcode such that the sequence information of the hits from the selection/screening can be easily tracked and decoded. To this end, biopolymers that can be conveniently sequenced by existing technologies, such as peptides and DNA, are among the popular candidates for molecular barcodes. Methods that expand the scope of the molecular barcodes to include other information-carrying synthetic polymers (e.g., the sequence-defined polymers for data storage discussed in Section 2.1), and polymer sequencing methods (e.g., mass spectrometry-based and nanopore-based sequencing techniques¹⁵⁷), will help to simplify the combinatorial library synthesis of novel sequence-controlled polymers and improve the efficiency of the selection/screening processes.