Proteins, deoxyribonucleic acids (DNAs), and polysaccharides are the three main natural biopolymers and are essential for biological systems and the maintenance of homeostasis. For example, glycopolymers, including polysaccharides and oligosaccharides, are used for lectin recognition, as components of the cytoskeleton, and for protein stabilization [1,2,3,4]. Proteins have numerous functions, including enzyme reactions related to metabolism, immune responses, and molecular recognition [5,6,7]. However, recreating their functions using artificial polymer materials has proven difficult, and this has led to natural biopolymers becoming a focus in polymer chemistry as well as the biological and medical fields. A growing number of bioinspired materials have been developed, and investigation into their medical applications is considered cutting-edge research [8,9,10,11,12,13]. One of the best examples is the drug delivery system (DDS), which is based on enhanced permeation and retention (EPR) effects via the self-assembly of polymers such as amphiphilic block copolymer-based micelles and vesicles [14,15,16,17,18]. To further advance bioinspired materials, including DDS, unnatural bioinspired polymers have become attractive targets for researchers in polymer chemistry and biomedical sciences [19,20,21,22,23,24,25].

This review focuses on recent advances related to unnatural biopolymers of unnatural oligosaccharides [26], protein–polymer conjugation [27], and protein stabilization by synthetic polymers [28]. To produce unnatural oligosaccharides, a newly designed bicyclic monomer (Glc(MeOx)) composed of glucosamine (GlcN) and 2-methyl-2-oxazoline (MeOx) has been utilized. A MeOx ring on the bicyclic monomer was selectively opened by a cationogen, inducing cationic ring-opening polymerization (CROP). In this polymerization, Glc(MeOx) was not polymerized via the classical MeOx mechanism, which is explained in greater detail in this review. The obtained oligo(Glc(MeOx)) has a backbone formed of N-1,2-glycosidic bonds, which cannot be prepared using natural biological systems. This means that this unnatural oligosaccharide can only be prepared using the original system developed for its production, and its important features are also discussed in this review. Protein–polymer conjugates are generated through the modification of proteins by synthetic polymers [29,30,31], and amphiphilic/fluorous methacrylate-based random copolymers were previously designed using living radical polymerization (LRP) [27, 28, 32, 33]. Owing to the hydrophilic poly(ethylene glycol) (PEG) side chains, random copolymers formed chain-folding structures in water and had low cytotoxicity. A lysozyme (Lyz) conjugate was successfully synthesized using amphiphilic/fluorous random copolymers. Furthermore, α-chymotrypsin and Lyz were stabilized by a random copolymer in 2H,3H-perfluoropentane (2HPFP), which is a fluorous solvent.

Unnatural sugar-based polymers

Glycopolymers have important roles in biological systems; however, their biological functions have not yet been completely elucidated. One of the reasons for this is that the synthesis of glycopolymers is difficult [34,35,36,37,38]. Consequently, to elucidate their unknown biological functions, we must first develop methods by which to artificially synthesize them [39, 40]. Furthermore, glycopolymer-based materials (glycomaterials) are also being developed for biomedical applications, as they are considered biocompatible [41,42,43]. The main synthesis pathways of glycopolymers are as follows: (1) the modification of natural polysaccharides such as pullulan and cellulose by polymer reactions [44, 45]; (2) enzymatic polymerizations [46]; and (3) the polymerization of sugar-based monomers [47,48,49,50,51,52,53,54,55,56] (Fig. 1). The advantages of polymer reactions are that the original biofunctions have already been clarified and that it is easy to design useful functional glycomaterials. However, one of the disadvantages is that the extracted natural glycopolymers must be purified to remove endogenous compounds such as proteins and other glycocompounds, which increases the cost of glycomaterials for biomedical applications. It is also difficult to control regioselective reactions when working with polymers. In contrast to this method, enzymatic polymerization and the polymerization of sugar-based monomers can generate both natural and unnatural glycopolymers. For example, enzymatic polymerization using a functionalized sugar monomer can afford regioselectively functionalized glycopolymers [46]. Scientists have reasoned that due to their natural sugar-based polymer design and the recently reported accelerated blood clearance (ABC) phenomenon, only glycomaterials will be suitable as alternative polymers with PEG [57, 58].

Fig. 1
figure 1

General methods used to prepare glycopolymers. Three main pathways are used in the preparation of glycopolymers: a polymer reaction [44, 71], b enzymatic polymerization [46], and c the polymerization of sugar-based monomers [26, 48,49,50,51,52,53,54,55,56]. Reference numbers have been provided for each example

From the standpoint of polymer design, biodegradability is an important factor for medical applications. It is easy to design biocompatible glycomaterials using sugar-based monomers; however, the effect of the hydrolyzed sugar units inside the body must be considered. For example, advanced glycation end products (AGEs) bind with the receptors for AGEs (RAGE) on macrophages, inducing proinflammatory cytokines and reactive oxygen species (ROS), such as superoxide, hydroxy radical, and hydroxy peroxide [59,60,61,62]. These glycative, endoplasmic reticulum (ER), and oxidative stresses are strongly associated with many diseases, such as type 2 diabetes [63], Alzheimer’s disease [64], and poor oocyte developmental competence [65], although the promise of chemotherapy by N-acetyl glucosamine (GlcNAc) has been reported [66]. Nonbiodegradable glycopolymers carrying unnatural N-1,2-glycosidic bonds have thus been designed to further investigate this prospect in the future.

Synthesis and characterization

Design of unnatural oligosaccharides

To synthesize unnatural glycopolymers, a new bicyclic monomer carrying the MeOx ring at the C1 and C2 positions of the GlcN ring (Glc(MeOx)) was designed after the hydroxy groups were protected by acetyl groups (Fig. 2a). The obtained monomer was polymerized with a pair of tert-butyl iodine (tBuI) molecules as a cationogen and GaCl3 as a Lewis acidic catalyst in 1,2-dichloroethane (DCE) at 40 °C. Up to 86% of Glc(MeOx) was consumed, and the size-exclusion chromatography (SEC) curve shifted to a higher molecular weight (Fig. 2b; the molecular weight of the peak top, Mp (max) = 960; the number-average molecular weight Mn = 800; the dispersity of the molecular weight, Đ(SEC) = 1.05). After purification, the obtained oligomer showed clear sequential peaks whose intervals corresponded to the molecular weight (MW) of the monomer (Fig. 2c; m/z = 329.11). These results showed the successful synthesis of oligo(Glc(MeOx)), which was the target unnatural oligosaccharide, and the 1H nuclear magnetic resonance (NMR) spectrum also supported successful synthesis [26]. Further characterization of the 1H NMR spectrum showed that the same number of α- and β-N-1,2-glycosidic bonds had been constructed without stereoselectivity (α/β ~ 1/1). This polymerization was not a living polymerization, as several side reactions occurred, although these side reactions have not yet been completely elucidated. Acetyl group deprotection was conducted by hydrazine in DMF at 25 °C to produce the deprotected oligo(Glc(MeOx)), which was confirmed by SEC, NMR, and ESI-MS (Fig. 2d, e).

Fig. 2
figure 2

Design of a new polymerization method by which to produce unnatural oligoaminosaccharides carrying N-1,2-glycosidic bonds via cationic ring-opening polymerization. a Polymerization of the sugar-based bicyclic monomer (Glc(MeOx)) to synthesize the unnatural oligoaminosaccharide with N-1,2-glycosidic bonds via cationic ring-opening polymerization. b, d SEC curves and c, e ESI-MS spectra of oligo(Glc(MeOx)) obtained using this original system b, c before and d, e after acetyl deprotection. Polymerization of f Glc(MeOx) and g MeOx in the absence of Lewis acid catalysts with MeOTf, MeI, tBuI, and CSA ((–)−10-camphorsulfonic acid). Reprinted with permission from [26]. Copyright [2019] American Chemical Society

Mechanism of polymerization

After the successful synthesis of oligo(Glc(MeOx)), the polymerization mechanism was investigated using the reaction of equimolar amounts of Glc(MeOx) and tBuI in the absence of Lewis acidic catalysts (Fig. 2f, g). It was hypothesized that this polymerization would proceed through the classical MeOx mechanism in the early stages of this research. With the cationic polymerization of MeOx, the intermediate iminoether cations are generally separated owing to their stability. Furthermore, the propagation reaction in the SN2 reaction opens the MeOx ring (Fig. 3a). Nevertheless, the iminoether cation of Glc(MeOx) was not obtained at all during the polymerization of Glc(MeOx), but the ring-opened compound was, as confirmed by 1H NMR [26]. DCE did not react with Glc(MeOx) in this reaction either. These results suggested that the SN2 reactions were minimal if they occurred at all in this system. Namely, these results suggested that this polymerization would proceed through the SN1 mechanism to produce the ring-opened compound (Fig. 3b). This occurs because the C1 in Glc(MeOx) is on the secondary carbon, resulting in fewer SN2 reactions. This explanation was further supported by the lower levels of stereoselectivity, which meant that the SN1 reaction mechanism could not control the ratio of α- and β-N-1,2-glycosidic bonds in the oligomers. The polymerization in the absence of Lewis acids further supported this proposed mechanism. In general, MeOTf (Tf, –SO2CF3) and MeI efficiently initiated the cationic polymerization of normal MeOx (Fig. 2g). In contrast, it was not MeOTf and MeI but tBuI that efficiently initiated the polymerization of Glc(MeOx), probably due to the reduction in free space around the N atom in Glc(MeOx) when compared with MeOx (Fig. 2f). In summary, a novel polymerization mechanism was proposed for the cationic ring-opening polymerization of Glc(MeOx), mainly via sequential SN1 reactions (Fig. 3b). Previously, the ring-opening polymerization mechanism for sugar-based monomers was considered to be an SN2 reaction-based polymerization inducing higher stereoselectivity by using the neighboring group effect (Fig. 3c) [48,49,50]. The differences in these polymerization mechanisms are of great interest and should be investigated further.

Fig. 3
figure 3

Proposed mechanism for the general propagation reactions to synthesize poly(MeOx) and saccharide-based polymers. Polymerization mechanism of a MeOx, b Glc(MeOx), and c another previously reported sugar-based monomer [48,49,50]. Reprinted with permission from [26]. Copyright [2019] American Chemical Society

Physical properties and structures

The physical properties and structures of oligo(Glc(MeOx)) were investigated after polymerization, starting with the secondary structures of oligo(Glc(MeOx)) in solution (Fig. 4). In DCE and tetrahydrofuran (THF), positive Cotton effects were observed in the circular dichroism (CD) spectra, and the intensity was much higher than that of pentaacetyl glucosamine (5AcGlcN) (Fig. 4a–c). The positive Cotton effects of the protected oligo(Glc(MeOx)) were dependent on the temperature, while those of 5AcGlcN were found to remain completely consistent across all temperatures. These results showed that the Cotton effect with 5AcGlcN was derived from the anomeric carbon, while that of oligo(Glc(MeOx)) resulted from the secondary structure of the oligomer. Furthermore, the Cotton effect of oligo(Glc(MeOx)) became negative in water after acetyl group deprotection, and the patterns were different from those of GlcNAc in water (Fig. 4d, e). These results also supported the idea that oligo(Glc(MeOx)) formed regular secondary structures in both organic and aqueous solvents and that the secondary structures were altered by the design of the protected groups.

Fig. 4
figure 4

Physical properties of the obtained oligo(Glc(MeOx)). Characterization of oligo(Glc(MeOx)) before and after acetyl deprotection by ae CD spectra, f XRD, g DSC, and h TG/DTA. In the CD measurements, the temperature was changed from 5 °C to 80 °C (5 °C (gray), 10 °C (dark blue), 20 °C (blue), 25 °C (green), 40 °C (light green), 60 °C (orange), and 80 °C (red); [oligo(Glc(MeOx))] = 1.0 mg/mL). In the DSC measurements, the temperature was increased from 40 °C to 250 °C and decreased to 40 °C (scanning rate = 5 °C/min). Reprinted with permission from [26]. Copyright [2019] American Chemical Society

The bulk properties of oligo(Glc(MeOx)) were also studied using X-ray diffraction (XRD), differential scanning calorimetry (DSC), and thermogravimetric/differential thermal analysis (TG/DTA). First, the bulk structures of oligo(Glc(MeOx)) before and after acyl group deprotection were observed by XRD (Fig. 4f). The diffraction patterns of GlcNAc clearly showed peaks from the crystalline structure. Almost all peaks disappeared after polymerization, and broad patterns were observed in both cases. In general, the amorphous polymers showed broad patterns in the small angle range owing to the lack of regular structures. However, oligo(Glc(MeOx)) showed two broad patterns before and after acyl group deprotection. The first and second diffraction angles shifted from 8.5° to 10° and 23° to 21°, respectively. Although further investigations will be required after increasing the MW of oligo(Glc(MeOx)), their bulk states were not completely amorphous; instead, the crystal structures had some regular orientations. Indeed, the exothermic peak of the deprotected oligo(Glc(MeOx)) was observed at approximately 206 °C in the DSC curve, and that of the protected oligo(Glc(MeOx)) was not (Fig. 4g). These results also support the idea that the bulk structures changed upon deprotection and that the broad and distorted exothermic peak also resulted from the reduction in regular structure. The thermal stability of oligo(Glc(MeOx)) was also investigated using TG/DTA under a N2 atmosphere (Fig. 4h). The endothermic peak of GlcNAc was absorbed at approximately 200 °C and decomposed from 210 °C. Compared with GlcNAc, the protected oligo(Glc(MeOx)) was found to decompose gradually. It is difficult to completely dry the deprotected oligo(Glc(MeOx)) owing to deliquescence, and weight loss in the TG curve was observed from the start due to dehydration. In summary, the oligomers were found to be slightly more stable than GlcNAc, but further investigations will be required after increasing their MWs.

Bioapplication: Cytotoxicity and biodegradability

To investigate the future bioapplications of oligo(Glc(MeOx)), in vitro cytotoxicity against HeLa cells was assessed using a WST-8 assay (Fig. 5a). In this assay, the concentration of the deprotected oligo(Glc(MeOx)) was changed from 0.01 to 10 mg/mL. HeLa cells were seeded at 5000 cells/well in a 96-well plate and cultured for 24 h at 37 °C. The medium was changed to a solution of the deprotected oligo(Glc(MeOx)) in DMEM, and the cells were further incubated for 24 h. After 2 h of reaction with WST-8, the absorbance at 450 nm was measured. The cell viability was close to 100% even under the highest concentration of the deprotected oligo(Glc(MeOx)). Therefore, the deprotected oligo(Glc(MeOx)) could potentially be investigated further for in vivo applications.

Fig. 5
figure 5

In vitro studies of oligo(Glc(MeOx)) for the investigation of cytotoxicity and biodegradability. a Cell viability of oligo(Glc(MeOx)) against HeLa cells characterized using the WST-8 assay (n = 6). The data are expressed as the mean ± standard error of the mean (SEM). b SEC curves of the deprotected oligo(Glc(MeOx)) before and after the hydrolysis reaction with chitinase in water. Reprinted with permission from [26]. Copyright [2019] American Chemical Society

Finally, the in vitro biodegradability of the deprotected oligo(Glc(MeOx)) was examined using chitinase, which cleaves the O-glycosidic bond (Fig. 5b). As mentioned, the N-1,2-glycosidic bond is impossible to prepare in biological systems and requires a specifically developed polymerization system. Owing to the unique glycosidic bonds, the deprotected oligo(Glc(MeOx)) was not decomposed by chitinase, as confirmed by SEC before and after the enzyme reaction in water at 80 °C for 3 h, as the SEC curves hardly changed. This is one of the unique aspects of oligo(Glc(MeOx)), indicating that it could potentially be used to produce new glycomaterials in the future.

Future issues regarding unnatural oligoaminosaccharides

A new polymerization method to prepare unnatural oligoaminosaccharides carrying unnatural N-1,2-glycosidic bonds has been developed. However, the system still has several issues associated with polymer synthesis. The first is that further development is required to enable living polymerization to control the MW of the unnatural oligo(Glc(MeOx)). The side reactions will be cleared in the next step to address this issue. Furthermore, controlling the side reactions is also required to help further elucidate the details of the polymerization mechanism. Second, the stereoconfiguration must also be controlled, which is important not only for polymer synthesis but also for bioapplications.

The in vivo toxicity of deprotected oligo(Glc(MeOx)) must also be discussed in detail prior to future bioapplications. The unnatural oligomer bases on the natural saccharide must be assessed to determine if they are truly biocompatible. For example, biomedicines such as antibodies and proteins have recently been developed for active targeting, but they sometimes exhibit severe adverse effects even though they are natural proteins [67]. In the case of biodegradable glycomaterials, the hydrolyzed sugar may affect endogenous metabolic systems. For example, it may be difficult to control blood sugar levels in patients with diabetes [22,23,24,25]. Consequently, the in vivo toxicity of oligo(Glc(MeOx)) must be investigated not only in healthy mice but also in disease models, such as those for diabetes, from the standpoint of glycative stress [59,60,61,62,63,64,65]. Second, sugars are generally used for molecular recognition in endogenous biological systems. After the achievement of controlled stereoconfiguration of oligo(Glc(MeOx)), the affinity of this oligomer for proteins and some receptors must be assessed in the next stage of research. In addition, the influence of saccharides on proteins must be further elucidated in the future by artificially designing the mixture and conjugates of saccharides and proteins because some diseases are closely related to both glycative and ER stresses, such as the examples given above.

Protein-based polymers

Given the aforementioned issues, proteins and glycopolymers are also important targets in medical applications. However, for medical applications, the methods by which natural proteins are collected, purified, and stabilized are crucial due to effects such as denaturation. Thus, protein technologies to enable extraction, stabilization, and storage while preventing negative effects must be developed to enable biological and clinical applications. To achieve these goals, the PEGylation of proteins was reported in the 1970s, which enabled protein stabilization, the suppression of immunogenicity, and enhanced blood circulation time [29]. After this initial breakthrough, technologies for the stabilization and storage of proteins developed rapidly due to protein–polymer conjugates and protein encapsulation using polymers [29,30,31, 68, 69]. For example, proteins conjugated with trehalose- and sodium 4-styrenesulfonate-based polymers were found to stabilize proteins from several stimuli [9, 56]. One of the general disadvantages regarding protein conjugation is the decrease in protein bioactivity. Protein encapsulation is one strategy to overcome this disadvantage, although cleavable protein conjugates have also been recently developed [70]. Indeed, pullulan functionalized by a small amount of cholesterol (CHP) forms nanogel structures by intermolecular associations, showing the biofunctionality of molecular chaperons [71], which have promise for treating ER stress-related diseases such as polycystic ovary syndrome [72].

Fluorous compounds have been used to develop a new type of protein conjugation and stabilization using methacrylate-based polymers [27, 28]. Fluorous compounds are immiscible with both hydrophilic and hydrophobic compounds [73] and do not disturb the inner hydrophobic interaction in the proteins. Therefore, water-soluble fluorous nanoparticles could potentially be a new type of protein conjugate and stabilizing agent. Amphiphilic random copolymers form polymer chain-folding nanoparticles in water and can change their folding compartments according to the outer environment [33]. The authors originally developed amphiphilic/fluorous random copolymers carrying PEG and fluorous side chains, which enabled reversible fluorous and hydrophilic PEG compartments in water and a fluorous solvent. In addition, fluorous copolymers could contribute to the advancement of protein–polymer phase separation and intercellular phase separation of proteins and related drugs [74,75,76].

Protein–polymer conjugation

Polymer design

To synthesize a new type of protein–polymer conjugation, amphiphilic/fluorous random copolymers carrying a carboxylic acid (P1), a disulfide pyridine (P2), or an activated ester (P3) group in a terminal end were designed by reversible addition–fragmentation chain transfer (RAFT) polymerization (Fig. 6a). Here, poly(ethylene glycol) methyl ether methacrylate (PEGMA) and 1H,1H,2H,2H-perfluorodecyl methacrylate (17FDeMA) were used to synthesize amphiphilic/fluorous random copolymers, and the target composition was set as PEGMA/17FDeMA = 70/30 and 60/40 [27]. RAFT polymerization afforded three random copolymers, confirmed by SEC and 1H and 19F NMR (Fig. 6a; Mn(SEC) = 98,300–118,000; Đ(SEC) = 1.71–2.10; PEGMA/17FDeMA ~ 70/35 or 60/50). These random copolymers mainly formed large multichain aggregating micelles whose RH values were 115–210 nm in H2O.

Fig. 6
figure 6

The original development of protein–polymer conjugation using methacrylate-based amphiphilic/fluorous random copolymers. a Design of amphiphilic/fluorous random copolymers using RAFT polymerization. The copolymers then formed self-folding and associating structures in water. Cytotoxicity of P1 against b NIH 3T3 cells and c HUVECs. d Preparation of protein–polymer conjugates with thiolated lysozyme (Lyz–SH) in aqueous solutions of P2. Reproduced from [27] with permission from the Royal Society of Chemistry. Copyright [2015] the Royal Society of Chemistry


Initially, the in vitro toxicity of P1 was investigated against NIH 3T3 mouse embryonic fibroblast cells and human umbilical vein endothelial cells (HUVECs) using a LIVE/DEAD assay (Fig. 6b, c). These cells were seeded into 48-well plates at a density of 6000 cells per well and preincubated for 24 h at 37 °C. The culture medium was exchanged with working media including P1 ([P1] = 0.1, 0.5, and 1.0 mg/mL), and these cells were further incubated for 48 h. They were then stained using the LIVE/DEAD regent and observed using a fluorescence microscope. P1 did not show cytotoxicity against NIH 3T3 cells or HUVECs, even at the highest concentration ([P1] = 1.0 mg/mL), owing primarily to the PEG layers on the nanoparticles of P1. Therefore, these amphiphilic/fluorous random copolymers could potentially be used in bioapplications.

Protein conjugation

Finally, protein conjugates were prepared using amphiphilic/fluorous random copolymers of P2 and P3 against lysozymes with a few SH groups (Lyz–SH, 4 SH groups/Lyz), and synthesis was evaluated using sodium dodecyl sulfate‒polyacrylamide gel electrophoresis (SDS‒PAGE) with Coomassie blue as a staining agent. Lyz–SH was simply mixed with the PBS solution of P2 or P3, and the solutions were kept at 4 °C for 24 h to prepare the conjugates. Only the conjugation of P2 was successfully achieved, although P2 formed chain-folding aggregating micelles in H2O (Fig. 6d). Due to the disulfide bond in the terminal end of P2, P2-SS-Lyz resulted from the reaction. The disulfide bond itself may be useful for biomedical applications, as endogenous glutathione (GSH) and ROS have the capability to cleave disulfide bonds inside the cells [69, 77]. P2-SS-Lyz could alter bioactivity using intercellular stimuli.

Protein stabilization

Polymer design

Advances that enable protein storage without denaturation are important and could help to promote protein-related research in biochemistry, biology, and biomedical fields. It is always preferable to avoid the use of toxic compounds to stabilize proteins after extraction. However, several strong toxic stabilizing agents must be used, as proteins are easily denatured in H2O by external stimuli. To address this issue, protein storage technology was developed that used a fluorous solvent with an amphiphilic/fluorous random copolymer (P4) (Fig. 7a) [28]. The random copolymer was prepared by PEGMA and 1H,1H,2H,2H-perfluorooctyl methacrylate (13FOMA) via ruthenium-catalyzed living radical polymerization (Ru-LRP) [78,79,80,81,82,83], and the successful synthesis was confirmed by SEC, 1H and 19F NMR (Fig. 7; Mn(SEC) = 68,200, Đ(SEC) = 1.3; Mn(NMR) = 97,400, m/nobsd. = 119/87; Mw(MALLS) = 177,000, multiangle laser light scattering (MALLS)). Owing to the PEG and fluorous side chains, P4 formed PEG compartments in 2HPFP and fluorous compartments in H2O, which was confirmed by 1H and 19F NMR and dynamic light scattering (DLS) (RH = 8 nm in 2HPFP and 6 nm in H2O) [28, 33].

Fig. 7
figure 7

Strategy for protein encapsulation in a fluorous solvent by utilizing the original amphiphilic/fluorous random copolymers. a Design of amphiphilic/fluorous random copolymers via ruthenium-catalyzed living radical polymerization (Ru-LRP), and P4 afforded reversible hydrophilic PEG and fluorous compartments in water and a fluorous solvent. b, c Protein encapsulation of lysozyme and α-chymotrypsin by P4. Reproduced from [28] with permission from the Royal Society of Chemistry. Copyright [2016] the Royal Society of Chemistry

Protein stabilization in a fluorous solvent

Protein encapsulation by P4 in 2HPFP was conducted as shown in Fig. 7b, c. The bulk mixture of Lyz and P4, which was obtained by lyophilization of an aqueous solution of P4 and Lyz, was solubilized by 2HPFP, and Lyz was extracted by water from a 2HPFP solution. The higher-order structures of Lyz before and after exposure to 2HPFP was characterized by CD spectra (Fig. 8a). The original solution of Lyz showed a negative Cotton effect from 200 to 240 nm (Fig. 8a, black line; λmax = 208 nm), and this Cotton effect remained almost the same even in the presence of P4 (Fig. 8a, blue line). The Cotton effect corresponded to the original effect after the extraction of Lyz by water from the 2HPFP solution (Fig. 8a, red line), indicating that the higher-order structure of Lyz remained even after exposure to 2HPFP due to the PEG compartmentation by P4.

Fig. 8
figure 8

Lysozyme (Lyz) and α-chymotrypsin were stabilized using an amphiphilic/fluorous random copolymer in 2H,3H-perfluoropentane (2HPFP) without denaturation. a CD spectra of the (black) original Lyz and (blue) Lyz in the presence of P4 and (red) Lyz after extraction from the 2HPFP solution using water. The red line is consistent with the black line, which indicates that Lyz was almost completely recovered from the 2HPFP solution using water extraction. b, c Characterization of the protein encapsulation of Lyz with P4 in 2HPFP by DLS and a representative TEM image. d, e Enzyme activity of Lyz after exposure to PBS, 2HPFP, or dichloromethane solutions in the presence or absence of P4 at 25 °C for 24 h. The activity was investigated using the hydrolysis of Micrococcus lysodeikticus in PBS, d the reactions were monitored by measuring the absorbance at 450 nm with UV/Vis measurements, and e the relative activity against the original Lyz was determined (***p < 0.001). f The activity of α-chymotrypsin was characterized by the hydrolysis of N-benzoyl-L-tyrosine-p-nitroanilide into p-nitroaniline in water at 37 °C. The reaction was monitored by observing the increase in absorbance at 410 nm for 250 s (ΔAbs/250). α-Chymotrypsin was stored in 2HPFP or water at 4 °C for 3, 5, or 24 h. Reproduced from [28] with permission from the Royal Society of Chemistry. Copyright [2016] the Royal Society of Chemistry

After the fluorous encapsulation of Lyz was achieved, the enzyme activity was investigated. Lyz encapsulated by P4 was incubated at 25 °C for 24 h (2HPFP/H2O = 999/1, v/v), and the folding structure was confirmed using DLS and TEM images (Fig. 8b, c). The Lyz was then extracted by phosphate buffer solution (PBS, pH ~7.0). Interestingly, there was no significant difference in the activity of Lyz after the extraction when compared with the original (Fig. 8d, e). In contrast, the enzyme activity significantly decreased after the exposure of Lyz against 2HPFP without P4 and against dichloromethane. These investigations showed that the formation of the PEG compartment in 2HPFP was important for protein stabilization in organic solvents. Furthermore, this PEG compartment constructed by P4 in 2HPFP also stabilized α-chymotrypsin (Fig. 8f). The enzyme activity of α-chymotrypsin was not changed after exposure to P4 in 2HPFP, while the α-chymotrypsin activity was reduced for 24 h in aqueous solution.

Future issues in protein–polymer conjugates and protein stabilization

The author focused on the synthesis of conjugation for methacrylate-based polymers and the stabilization of protein in a fluorous solvent. As mentioned, protein stabilization is one of the key technologies required for the development of protein biomedical applications. After the administration of proteins into the body, hydrolysis by endogenous enzymes is unavoidable. In addition, the controlled switching of enzyme activity is important to suppress the severe adverse effects of biomedicines such as antibodies [67]. Protein–polymer conjugation and protein stabilization using synthetic polymers are promising methods to address the existing issues and must be developed for use in future biomedical technologies.


In this focus review, recent developments in the field of unnatural biopolymers that involve unnatural oligoaminosaccharide and protein–polymer conjugation have been discussed. In the saccharide project described, a newly designed polymerization method and unnatural oligo(Glc(MeOx)) with N-1,2-glycosidic bonds were reported, and the oligomer can be synthesized only by using the author’s original system. A new polymerization mechanism has thus been proposed that is different from that using normal 2-oxazolines. Furthermore, the successful synthesis of protein–polymer conjugates and stabilized proteins in a fluorous solvent by amphiphilic/fluorous methacylate-based random copolymers was also discussed. Both achievements have issues that will require further investigation, and this research is ongoing with the aim of improving their use in biomedical applications.