Protein-Stabilizing Effect of Amphiphilic Block Copolymers with a Tertiary Sulfonium-Containing Zwitterionic Segment.

Tertiary sulfonium-containing zwitterionic block copolymers consisting of N-acryloyl-l-methionine methyl sulfonium salt (A-Met(S+)-OH) and n-butyl acrylate (BA) were newly synthesized to develop a novel protein stabilizer. The zwitterionic block copolymers were prepared by reversible addition-fragmentation chain-transfer (RAFT) polymerization of BA using a hydrophilic macro-chain-transfer agent (CTA) obtained from N-acryloyl-l-methionine (A-Met-OH) and subsequent postmodification. RAFT polymerization of A-Met-OH using poly(BA) macro-CTA, followed by postmodification, also afforded the target poly(A-Met(S+)-OH)-b-poly(BA). The block copolymers stabilized horseradish peroxidase (HRP) during storage at 37 °C for 5 days, and the protein-stabilizing effect was enhanced with increase in the A-Met(S+)-OH content. In particular, the block copolymer with ∼85% A-Met(S+)-OH content showed a significant protein-stabilizing effect at a temperature (37 °C) higher than the room temperature, which is highly desirable for practical and industrial applications. The addition of sucrose into the block copolymer-protein solution led to a considerable increase in the HRP activity under the same conditions. Excellent alkaline phosphatase stabilization at 37 °C for 12 days was also achieved using the block copolymers. The zwitterionic block copolymers with the optimal hydrophilic/hydrophobic balance were found to serve as efficient protein-stabilizing agents, in comparison with the corresponding homopolymer and random copolymers. Dynamic light scattering, zeta potential, transmission electron microscopy, and circular dichroism measurements revealed that the zwitterionic block copolymer stabilizes an enzyme by wrapping with a slight change in the size, whereas the secondary and ordered structures of the enzyme are maintained.

Introduction

The past few decades have seen a significant interest in zwitterionic copolymers, owing to their unique and diverse features, especially their excellent protein resistance.[1−4] These zwitterionic polymers contain oppositely charged ionic groups on the same molecule.[5] Because of the structural similarity between the zwitterionic group and the cell membrane, zwitterionic polymers have also been employed as protein models and biointerfaces for medical,[6−8] diagnostic, and biotechnological applications.[9,10] Zwitterionic polymers, also known as betaine polymers, are generally categorized into phosphobetaine,[11] sulfobetaine,[12−15] and carboxybetaine.[16−20] In most cases, the ammonium- and nitrogen-containing ionizable groups have been used as cationic components of the zwitterionic polymers.

Recently, a novel zwitterionic polymer containing tertiary sulfonium cationic groups has emerged, which can be regarded as one of the most promising candidates for various applications.[21,22] A variety of sulfonium-based zwitterionic polymers have been recently developed, which comprise sulfothetins (i.e., sulfonium sulfonates)[23] and carboxythetins (i.e., sulfonium carboxylates).[24,25] Very recently, we reported the controlled synthesis of a tertiary sulfonium-containing zwitterionic polymer by reversible addition–fragmentation chain-transfer (RAFT) polymerization of N-acryloyl-l-methionine (A-Met-OH), followed by postmodification.[26] RAFT polymerization was selected, as it enables controlled syntheses of various zwitterionic polymers from zwitterionic and other functional monomers without the protection chemistry or postpolymerization modification.[5,27] RAFT is the most versatile with respect to the functional monomer and the reaction medium, which leads to the development of a variety of polymeric architectures with functional groups. Additionally, ease of scale-up of the reaction process and no possible deactivation of proteins and biologically active substances by metal residues make it a promising technique to produce well-defined polymers for protein stabilization. The resulting zwitterionic component, with a tertiary sulfonium group and a carboxylic acid moiety in the monomer unit (N-acryloyl-l-methionine methyl sulfonium salt: A-Met(S+)-OH, which belongs to carboxythetins), has a structure similar to the protein-protecting agent dimethylsulfoniopropionate (DMSP), which is released by marine algae.[28−30] DMSP, with the general formula (CH3)2S+CH2CH2COO–, is known to have antioxidant activity as well as the ability to protect cells from UV damage.[31,32] Our previous investigation indicated that DMSP-mimicking zwitterionic poly(A-Met(S+)-OH) has low cytotoxicity and stabilizes horseradish peroxidase (HRP), peroxidase-labeled mouse antibody immunoglobulin G (IgG-HRP), and alkaline phosphatase (ALP) during storage (at 4 °C for HRP and IgG-HRP and at 37 °C for ALP) and freeze–thaw cycles, even at low concentrations.[26] Amphiphilic random copolymers consisting of A-Met(S+)-OH and n-butyl acrylate (BA), obtained via RAFT copolymerization and subsequent postmodification, exhibited excellent protein stabilization properties. However, the protein-stabilizing effect of the tertiary sulfonium-containing amphiphilic block copolymer has not yet been investigated.

In this report, we have tested the protein-stabilizing effect of novel tertiary sulfonium-containing zwitterionic block copolymers, poly(A-Met(S+)-OH)-b-poly(BA)s, which present a novel paradigm, where decrease in protein activity can be suppressed by controlling the hydrophilic/hydrophobic (zwitterionic A-Met(S+)-OH/nonionic BA) balance and monomer sequence (block or random). To the best of our knowledge, this is the first report on the RAFT synthesis of sulfonium-containing zwitterionic block copolymers showing excellent protein stabilization effects. Practical application of proteins and antibodies as pharmaceuticals or detection of highly sensitive proteins using antibodies requires the development of a variety of protein-stabilizing macromolecules. Several polymer-like excipients, such as bovine serum albumin (BSA),[33] polyethylene glycol and polypropylene glycol,[34] glycopolymers,[35] and 2-methacryloyloxyethyl phosphorylcholine (MPC)-containing polymers,[36] have been employed as additives to prevent the aggregation and denaturation of proteins. A variety of polymers have been developed via grafting of a polymer chain, via in situ polymerization of a cross-linked shell, or by assembling a preformed polymer to extend the enzyme shelf life.[37] Enzymes can be shielded from aggressive environments by using these polymers, thus delaying denaturation and increasing control of the immediate environment around the enzyme and consequently resulting in a possible control of substrate diffusion. However, the development of novel, more effective, and nontoxic synthetic stabilizers, which preserve the structure and/or activities of enzymes for long periods of time under physiological and more severe conditions, is still required.

Block copolymer micelles are an important class of aqueous colloids that find several applications in the biomedical field, such as formulation of lipophilic drugs in cancer treatments and delivery of oligonucleotides in gene therapy.[38,39] Several amphiphilic block copolymers have been employed to prevent the aggregation of unfolded proteins.[40] For example, triblock copolymers and tetrafunctional block copolymers composed of hydrophobic poly(propylene oxide) and hydrophilic poly(ethylene oxide) chains were reported to reduce the aggregation of the unfolded lysozyme, signifying the importance of a match in size between the hydrophobic region of the denatured protein and that of the amphiphilic copolymer.[41] Similarly, amphiphilic copolymers based on propylene oxide and ethylene oxide were also effective in preventing the aggregation of denatured and reduced lysozyme, and these amphiphilic copolymers can be used as a synthetic chaperone in the innate molecular repair mechanism of cells.[42]

Zwitterionic block copolymers, which are central to the development in this field, can be classified into two broad categories: polyampholytes, in which cationic and anionic groups are covalently linked, and polybetaines, which have oppositely charged groups in the same monomer unit.[5,27] The latter block copolymers composed of zwitterionic (polybetaine) and nonionic groups belong to a group of amphiphilic block copolymers that self-assemble into diverse morphologies because of incompatibility between at least one of the building blocks and the solvent. Diverse micellar constructs with a compact hydrophobic domain and a swollen hydrophilic shell are formed in aqueous media and exhibit characteristic morphologies and properties in response to the external environment.[11,43−45] Several zwitterionic block copolymers containing covalently linked zwitterionic and hydrophobic segments have been employed as protein-stabilizing macromolecules. For example, Ishihara et al. reported that the stabilizing effect of the protein was greatly improved when a block copolymer composed of MPC and butyl methacrylate was modified on the enzyme surface.[36] They also reported the synthesis and protein stabilization properties of MPC-based random copolymers composed of hydrophobic styrene[46] and n-dodecyl methacrylate.[47]

In this work, a novel class of zwitterion-based block copolymer micelles has been produced, and a method to control the comonomer composition, assembled structure, and protein-stabilizing ability has been demonstrated. This has been achieved by the synthesis and fine-tuning of zwitterionic block copolymers obtained by RAFT polymerization. While the hydrophobic group usually serves as the core-forming unit, the second block of the copolymer consists of a hydrophilic polymer that is responsible for the stabilization of the block copolymer micelles when located in the micellar shell. Limited solubility in common organic solvents creates challenges in producing large libraries of amphiphilic copolymers based on zwitterionic monomers and in controlling the properties of the final synthesized product. In other words, the high hydrophilicity of the zwitterionic polymers makes the last step of the block copolymer synthesis critical because of the high difference in hydrophilicity (and lipophilicity) between the two blocks. In this study, the zwitterionic block copolymers were synthesized via a two-step procedure, which consists of the RAFT polymerization of BA using dithiocarbamate-terminated poly(A-Met-OH) and subsequent modification of the methyl thioether moiety (Scheme a). In this approach, the initial synthesis of the amphiphilic block copolymer with the anionic (carboxylic acid-containing) poly(A-Met-OH) segment can be regarded as a precursor to the targeted zwitterionic segment and allows higher solubility in the reaction mixture during polymerization. Poly(A-Met-OH) acted as an anionic macro-chain-transfer agent (CTA), thereby minimizing the difficulty in synthesis in terms of reactivity and solubility. Rapid conversion of the poly(A-Met-OH) macro-CTA to the block copolymer is required, without unfavorable side reactions, allowing the chains to initiate the formation of the second poly(BA) block at approximately the same time[48,49] to obtain block copolymers with the as-designed structure and low dispersity. BA was selected as a hydrophobic monomer for the synthesis because of the various advantageous properties of BA-containing copolymers, such as excellent biocompatibility as well as high flexibility, stability, mechanical properties, and good adhesion. Hence, BA is frequently employed as a comonomer in biocompatible soft materials, including gels,[50−52] polymer networks,[53,54] and random copolymers.[55,56] The zwitterionic–hydrophobic block copolymer was also synthesized by RAFT polymerization of the unprotected amino acid-based monomer (A-Met-OH) from poly(BA) macro-CTA, followed by the same modification of the methyl thioether moiety (Scheme b). In our system, both acrylate and acrylamide belong to the conjugated monomer group, which may support efficient block formation. Limited solubility of the zwitterionic poly(A-Met(S+)-OH) in organic solvents is a challenge, especially when attempting the addition of the hydrophobic poly(BA) block. Distinct from the zwitterionic homopolymer, poly(A-Met(S+)-OH), the carboxylic acid-containing poly(A-Met-OH) was soluble in various organic solvents, such as methanol, ethanol, and dimethylformamide (DMF), which enables the synthesis of targeted zwitterionic block copolymers. In both systems, a positively charged sulfonium group in the poly(A-Met(S+)-OH) segment can be recognized as an inherently quenched ionic species regardless of the pH range, whereas the carboxylic acid is subject to protonation/deprotonation equilibria depending on the pH value. A systematic study of the effect of varying block copolymer compositions is presented, which enables the identification of a suitable zwitterionic–nonionic combination to achieve high protein stabilization. The effect of the comonomer sequence (block or random) on protein stability was also investigated. In the system developed in this study, specific interactions between proteins and polymers can be tuned by modifying two parameters, the hydrophilic/hydrophobic balance and the monomer sequence, which correspond to the location and distribution of a tertiary sulfonium-containing zwitterionic A-Met(S+)-OH unit in the copolymers and hence an efficient approach in protecting a targeted protein.

Scheme 1

(a) Synthesis of Amphiphilic Block Copolymers by (a) RAFT Polymerization of BA Using Poly(A-Met-OH) as a Hydrophilic Macro-CTA, Followed by Postmodification and (b) RAFT Polymerization of N-Acryloyl-l-methionine (A-Met-OH) Using Poly(BA) as a Hydrophobic Macro-CTA, Followed by Postmodification

Results and Discussion

Synthesis of Zwitterionic Block Copolymers, Poly(A-Met(S+)-OH)-b-poly(BA)

In this study, we report the controlled synthesis of novel sulfonium-containing zwitterionic block copolymers, with unique amphiphilic (zwitterionic–nonionic) properties, by RAFT polymerization, and their self-assembled structures, as well as their protein-stabilizing effects. The zwitterionic block copolymers, poly(A-Met(S+)-OH)-b-poly(BA)s, were initially synthesized by RAFT polymerization of BA using the dithiocarbamate-terminated poly(A-Met-OH), followed by postmodification (Scheme a). Poly(A-Met-OH) macro-CTA was initially synthesized by RAFT polymerization of A-Met-OH using a dithiocarbamate-type CTA and 2,2′-azobis(isobutyronitrile) (AIBN) as the initiator in DMF at 65 °C according to a previously reported procedure.[26] Following polymerization at a proportion of [M]/[CTA]/[AIBN] = 100/2/1 for 3 h, the reaction mixture was purified by reprecipitation in a large excess of methyl tert-butyl ether (MTBE), and the purified poly(A-Met-OH) obtained was a yellow solid. The size exclusion chromatography (SEC) trace of the methylated sample, poly(A-Met-OMe), indicated a narrow molar mass distribution (Mw/Mn = 1.23). The molar mass of the poly(A-Met-OH) (Mn,NMR was 12 000) was estimated from the relative integration of the 1H NMR signals for the polymer backbone protons (−CH2CH2–S– unit at 2.5–2.8 ppm) and the end-group protons (benzyl unit at 7.0–7.3 ppm), as shown in Figure S1a (Supporting Information). The dithiocarbamate-terminated poly(A-Met-OH) having the carboxylic acid moiety in the monomer unit was employed as a hydrophilic macro-CTA (Mw/Mn = 1.23, Mn,NMR = 12 000, Mn,SEC = 8200) for the synthesis of the amphiphilic block copolymers.

Subsequently, poly(A-Met-OH)-b-poly(BA)s were synthesized by RAFT polymerization of BA using the poly(A-Met-OH) macro-CTA in ethanol, which is a good solvent for both components. In this study, the polymerization conditions were adjusted to keep the system homogeneous and to avoid unfavorable side reactions. In order to manipulate the comonomer composition, the [M]0/[macro-CTA]0 molar ratio was adjusted from 5 to 115 under suitable conditions. The results are summarized in Table . Polymerization proceeded in a homogeneous manner when conducted in ethanol ([M] = 1.0 M) in the presence of AIBN as the initiator at 65 °C. After the reaction mixture was reprecipitated into MTBE, poly(A-Met-OH)-b-poly(BA)s were obtained in reasonable yields (37–77%). Because the CTA-to-initiator ratio was fixed to [macro-CTA]/[I] = 1/0.5 in this system, the higher [M]/[macro-CTA] ratio leads to the decrease in the initiator concentration, resulting in the decrease in the polymer yield. In contrast, significant increase in the poly(BA) chain length, relative to that of poly(A-Met-OH) macro-CTA, may induce higher chain flexibility to increase the diffusion rate of the propagating radical, owing to better solubility of the poly(BA) chain, compared to that of the poly(A-Met-OH), in ethanol. Both factors, the initiator concentration and radical diffusion rate, may contribute to the polymer yields.

Table 1

Synthesis of the Amphiphilic Block Copolymer, Poly(A-Met(S+)-OH)-b-poly(BA)s, in Ethanol at 65 °C for 20 ha

				Mn (×103)
entry	macro-CTA	[M]/[macro-CTA]	conv.b/yieldc (%)	Theod	NMRb	SECe	Mw/Mne	Comp.b A-Met(S+)-OH/BA
BP1f	poly(A-Met-OH)	5	77/72	13.3	13.1	8.7	1.36	96/4
BP2f	poly(A-Met-OH)	15	81/78	14.3	14.3	10.0	1.33	83/17
BP3f	poly(A-Met-OH)	21	41/37	13.9	14.1	8.6	1.16	85/15
BP4f	poly(A-Met-OH)	115	78/77	24.3	29.7	11.0	1.41	31/69
BP5g	poly(BA)	50	100/95	15.9	16.8	18.3	1.54	67/33

Polymerization with AIBN as an initiator ([CTA]/[AIBN] = 2) at 65 °C for 20 h in EtOH ([M] = 1.0 M), where A-Met-OH = N-acryloyl-l-methionine, A-Met(S+)-OH = N-acryloyl-l-methionine methyl sulfonium salt, and AIBN = 2,2′-azobis(isobutyronitrile).

Calculated by 1H NMR in CD3OD.

MTBE-insoluble part.

For BP1–4; Mn,theory = (Mw of BA) × [BA]0/[macro-CTA]0 × conv. + (calculated Mn of poly(A-Met(S+)-OH) as macro-CTA), For BP5; Mn,theory = (Mw of A-Met(S+)-OH) × [A-Met-OH]0/[macro-CTA]0 × conv. + (Mn of macro-CTA).

Measured by SEC using MMA standards in DMF containing 1 g/L LiBr.

Poly(A-Met-OH) (Mn,NMR = 12 000, Mn = 8200, and Mw/Mn = 1.23 in the methylated form) was used as macro-CTA (calculated Mn of poly(A-Met(S+)-OH) = 12,800, DP = 58.5).

Poly(BA) (Mn,NMR = 4100) was used as macro-CTA.

The block copolymer composition was determined using 1H NMR spectroscopy by a comparison of the peaks associated with the two comonomers, that is, the peak at 2.5–2.8 ppm attributed to the S–CH2– proton (2H) of the A-Met-OH unit and the peak at 0.9–1.0 ppm corresponding to the CH2–CH3 peak protons (3H) of the BA unit (Figure a). The BA content in the block copolymers varies from 4 to 69 mol % depending on the comonomer feed ratio. Similar to the homopolymer, the resulting poly(A-Met-OH)-b-poly(BA)s required derivatization prior to SEC analysis in DMF containing 1 g/L LiBr as the eluent. The carboxylic acid units in the poly(A-Met-OH) segment of the amphiphilic block copolymers were converted into methyl ester forms, poly(A-Met-OMe)-b-poly(BA)s, by TMS-diazomethane. As shown in Figure , SEC analysis of the resulting poly(A-Met-OMe)-b-poly(BA)s revealed a reasonable shift in the unimodal SEC peak toward a higher molar mass (BP4; Mn,SEC = 11 000), with an increase in [M]0/[CTA]0 ratio while maintaining low dispersity. The molecular weights of the resulting poly(A-Met-OH)-b-poly(BA)s determined by 1H NMR were in reasonable agreement with the theoretical values calculated from the conversion and feed ratio of the comonomer. The molecular weights of the poly(A-Met-OH)-b-poly(BA) (BP4; Mn,NMR was 29 000, which corresponds to 29 700 in the zwitterionic form) were estimated from the relative integration of the 1H NMR signals for the peaks corresponding to the two comonomers (the peak at 2.5–2.8 ppm attributed to the S–CH2– proton and the peak at 1.0 ppm corresponding to the −CH2-CH3 protons). Note that SEC analysis of these ionic–nonionic block copolymers was challenging because these species are prone to hydrophilic and electrostatic interactions with the stationary phase. In this study, the methylated block copolymers having relatively low BA contents showed unimodal SEC traces (Figures and S2, Supporting Information) with reasonable molar mass, which were measured in DMF (1 g/L LiBr) using poly(methyl methacrylate) standards for calibration. However, the block copolymers having relatively high BA contents (BP4 and BP5) may lead to difficulty in accurate determination of the molar mass by SEC measurements. Another possible reason for the discrepancy between Mn,NMR and Mn,SEC as well as the theoretical values is the occurrence of partial terminations and unfavorable side reactions at the last stage of the polymerization. Nevertheless, these results clearly demonstrate that the chain extension of the poly(A-Met-OH) macro-CTA with BA was well-controlled in ethanol under suitable conditions, by which the composition of each segment and molar mass of poly(A-Met-OH)-b-poly(BA)s could be tuned by adjusting the monomer-to-macro-CTA ratio in the feed.

Figure 1

1H NMR spectra of (a) poly(A-Met-OH)-b-poly(BA) in CD3OD, (b) poly(A-Met(S+)-OH)-b-poly(BA) (BP4) in CD3OD, (c) poly(A-Met(S+)-OH)-b-poly(BA) (BP4) in CD3OD + 20% D2O, and (d) poly(A-Met(S+)-OH)-b-poly(BA) (BP4) in D2O.

Figure 2

SEC traces of (a) poly(A-Met-OMe)-b-poly(BA)s (BP2 and BP4) obtained from poly(A-Met-OH) macro-CTA and (b) poly(BA)-b-poly(A-Met-OMe) (BP5) obtained from poly(BA) macro-CTA. See Table for detailed polymerization conditions.

In a previous report, we demonstrated that both the structure of the chain-end groups and polydispersity significantly influence the protein stabilization ability of the zwitterionic polymer, poly(A-Met(S+)-OH).[26] As shown in Figure , the chain end groups of poly(A-Met-OH)-b-poly(BA)s were not detected clearly, suggesting the possibility to change the chain-end structures by partial terminations and unfavorable side reactions. Distinct from the homopolymer, nevertheless the effect of the end groups is not crucial in the case of the block copolymers because the hydrophobic poly(BA) segment should affect predominantly the self-assembled structures and protein stabilization.

The carboxylic acid-containing amphiphilic block copolymers, poly(A-Met-OH)-b-poly(BA)s, were modified using iodomethane to afford zwitterionic block copolymers consisting of hydrophilic poly(A-Met(S+)-OH) and hydrophobic poly(BA) segments. The resulting poly(A-Met(S+)-OH)-b-poly(BA)s were purified by dialysis against methanol and water, followed by lyophilization. Note that the anionic/nonionic poly(A-Met-OH)-b-poly(BA)s were purified by the reprecipitation into MTBE, whereas the purification of the zwitterionic–nonionic poly(A-Met(S+)-OH)-b-poly(BA)s was conducted by dialysis, which is mainly due to the difference in the solubility between the block copolymers composed of anionic poly(A-Met-OH) and zwitterionic poly(A-Met(S+)-OH) (see Tables S1 and S2, Supporting Information). The formation of poly(A-Met(S+)-OH)-b-poly(BA) was confirmed using 1H NMR spectroscopy. A larger downfield shift of the signal corresponding to the methyl group of the pendant thioether (δ = 2.1 ppm) in poly(A-Met-OH)-b-poly(BA) compared to that of the newly formed sulfonium species (δ = 2.9–3.1 ppm) in poly(A-Met(S+)-OH)-b-poly(BA) was observed in CD3OD (Figure b). Similar to the case of the zwitterionic homopolymer,[26] the almost quantitative modification was confirmed by 1H NMR spectroscopy (Figure c) by comparing the integration of the methylene resonance at 3.4–3.8 ppm (CH2–CH2–S(CH3)2) with the intensity of the methine resonances at 4.3–4.7 ppm (NH–CH–COOH), as well as the peak at 4.3–4.7 ppm (NH–CH–COOH) and the peak at 3.0–3.2 ppm (CH2–S(CH3)2). In all cases (BP1–BP3 and BP5), the degree of the modification to Met(S+) in each copolymer was more than 99%, which was confirmed by 1H NMR. The 1H NMR spectrum of poly(A-Met(S+)-OH)-b-poly(BA) (BP4) measured in D2O is shown in Figure d. The measurement solution was resuspended in D2O, and the peak (E) derived from BA disappeared, in addition to the significant decrease in the integrated value of the methyl protons of the BA unit (peak H). Hence, it is reasonable to consider that poly(A-Met(S+)-OH)-b-poly(BA) forms micelles in water, in which the water-insoluble poly(BA) segment is located inside the core, whereas the water-soluble poly(A-Met(S+)-OH) segment forms the shell part that forms an interface with water.

After postmodification, the zwitterionic functionality of the synthesized block copolymers is quantitatively formed, which leads to the increase in the water solubility derived from effective micelle formation. Similar to the zwitterionic homopolymer poly(A-Met(S+)-OH), poly(A-Met(S+)-OH)-b-poly(BA)s with a high zwitterionic content (A-Met(S+)-OH content ≥ 83%) were soluble in water, and the block copolymers with lower A-Met(S+)-OH content (≤67%) were difficult to dissolve in water (Table S2, Supporting Information). This is in accordance with a general tendency that the overall solubility of amphiphilic block copolymers can be controlled by changing the respective blocks in the composition. The zwitterionic block copolymers were found to be soluble in CH3OH + 20% H2O, independent of the comonomer composition. The tendency was apparently distinct from that of the anionic (carboxylic acid-containing)–nonionic poly(A-Met-OH)-b-poly(BA)s, which exhibited good solubility in acetone, methanol, ethanol, and DMF, while being insoluble in water. Note that the anionic block copolymers having relatively high poly(A-Met-OH) contents (≥67%) were soluble in the CH3OH/H2O mixed solvent (Table S1, Supporting Information).

We also attempted to synthesize the zwitterionic block copolymers by RAFT polymerization of A-Met-OH using a hydrophobic poly(BA) as a macro-CTA (Scheme b). When designing a block copolymer by RAFT polymerization, the order of preparation of the blocks determines the final position of the RAFT end group. In the second approach, a hydrophobic poly(BA) macro-CTA was prepared first, followed by chain extension with the hydrophilic A-Met-OH as a second monomer (Scheme b). In theory, the reverse order would be possible because both building blocks are based on the conjugated monomers, acrylate and acrylamide. A hydrophobic macro-CTA, poly(BA), was synthesized by RAFT polymerization of BA using a dithiocarbamate-type CTA with AIBN as the initiator in DMF at 65 °C for 2 h. After polymerization, the reaction mixture was dialyzed against acetone for 2 days [molecular weight cutoff (MWCO): 1000 Da]. The 1H NMR signals for the polymer backbone protons (O–CH2CH2– unit at 3.9–4.1 ppm) and the end group protons (benzyl unit at 7.0–7.3 ppm) were employed for the determination of the molar mass (Mn,NMR = 4100, Figure S3 in the Supporting Information). Poly(BA)-b-poly(A-Met(S+)-OH) was synthesized by RAFT polymerization of A-Met-OH using the dithiocarbamate-terminated poly(BA) in ethanol with AIBN as the initiator at 65 °C. As shown in Figure b, a clear shift of the unimodal SEC peak from the macro-CTA was observed for the resulting poly(BA)-b-poly(A-Met-OMe). The 1H NMR spectrum of the block copolymer (BP5) showed peaks corresponding to the two comonomers (BA and A-Met(S+)-OH, Figure S3, Supporting Information), suggesting the formation of the targeted block copolymer with intermediate BA content (33%).

The glass-transition points (Tg) of the poly(A-Met(S+)-OH)-b-poly(BA)s were determined by differential scanning calorimetry (DSC) measurements (Figure S4). The observed Tg values were −51.2 °C for BP4 and -52.9 °C for BP5 (BA content = 69 and 33%, respectively), which are attributed to that of the poly(BA) segment (Tg of poly(BA) = −55 °C). In contrast, the block copolymers having relatively low BA content exhibited no detectable transition temperature. These results indicate the formation of block copolymers having different poly(BA) chain lengths by RAFT polymerization.

Assembled Structures of Poly(A-Met(S+)-OH)-b-poly(BA)s in Aqueous Solutions

Assembled structures of the zwitterionic block copolymers with 31–96% A-Met(S+)-OH contents were investigated in aqueous solution under various conditions. Initially, the property of dissolution in water (pH = 7), which is a good solvent for poly(A-Met(S+)-OH) of poly(A-Met(S+)-OH)-b-poly(BA), was characterized using dynamic light scattering (DLS). While water is a good solvent for zwitterionic block copolymers with a short hydrophobic segment, the zwitterionic block copolymers with longer hydrophobic segments (BA content ≥ 33%) can only be dissolved in a mixed water–methanol solvent. Because of the amphiphilic properties of the block copolymers, the synthesized materials can directly self-assemble into nanoparticles if dissolved in an organic solvent and then added dropwise into an aqueous medium. A concentrated solution of the block copolymers in CH3OH + 20% H2O, which forms a good solvent for both segments, was gradually dispersed by the slow addition of water. Samples were filtered (0.8 μm), followed by the measurement at 25 °C. As shown in Figure S5 and Table S3 (Supporting Information), poly(A-Met(S+)-OH)-b-poly(BA) with low BA content (4%) showed a unimodal distribution of hydrodynamic diameters in the selected solvent, and the relatively narrow distribution (Dh = 19.8 nm) probably indicates the existence of spherical micelles consisting of a relatively hydrophobic core of poly(BA) and a hydrophilic shell of poly(A-Met(S+)-OH) in neutral water (pH = 7). In the case of the zwitterionic block copolymers, the surface of the assembled structure is expected to consist of an equal number of positive and negative charges, resulting in a zeta potential close to 0. The zeta potential of BP1 in water was −7.2 mV at pH = 7.4 (Figure S6, Supporting Information), suggesting that poly(A-Met(S+)-OH)-b-poly(BA) (BP1) is electrically neutral with no significant deviation of charge under the conditions employed. As expected, the size of the micelles increased with increasing hydrophobic poly(BA) content. As shown in Figure , the relatively narrow hydrodynamic diameter distribution (Dh = 22.4 nm), which may correspond to the uniform micelles consisting of a poly(BA) core and a poly(A-Met(S+)-OH) shell, was still detected for the block copolymer with a 15% BA unit, whereas a significant increase in the size (Dh = 146 nm) was observed in the block copolymer with a higher BA content (69%), which is probably attributed to the aggregated micelles.

Figure 3

Z-averaged hydrodynamic diameter distributions of block copolymers, poly(A-Met(S+)-OH)-b-poly(BA)s (Dh, as determined by DLS): (a) BP3 and (b) BP4. The A-Met(S+)-OH/BA composition = (a) 85/15 and (b) 31/69, respectively. (c) Intensity-averaged hydrodynamic diameter distributions (Dh) against the composition of the zwitterionic A-Met(S+)-OH unit of the block and random copolymers.

In comparison, the zwitterionic random copolymers with 32–86% A-Met(S+)-OH contents were investigated in aqueous solution under the same conditions. We employed the random copolymers, poly(A-Met(S+)-OH-co-BA)s, which were synthesized in our previous work[26] via copolymerization of A-Met-OH and BA, followed by methylation using iodomethane (Table S4, Supporting Information). Owing to the polar nature of the A-Met(S+)-OH unit, the synthesized amphiphilic random copolymers were soluble in water depending on the comonomer composition and soluble in water–methanol mixed solvent independent of composition.

In both block and random copolymers with A-Met(S+)-OH units, poly(BA) enriches the hydrophobic property of its derived copolymers to enable better compatibility with the targeted proteins. As shown in Figures c and S7 (Supporting Information), the hydrodynamic diameter distributions (Dh) were 97 and 111 nm for the random copolymers with a higher BA content (50 and 68%), which may be ascribed to the aggregated micelles. In contrast, random copolymers with a higher A-Met(S+)-OH content (86 and 68%) showed no detectable DLS peaks, suggesting no assembled structures under the conditions (Table S4, Supporting Information). In both block and random copolymers with a relatively high BA content (>50%), relatively large species were formed in aqueous solutions, which are closely related to the protein stabilization properties described in the next section.

Protein Stabilization Test (Accelerated Degradation Test)

We have previously reported that hydrophobicity in polymers greatly contributes to the protein-stabilizing effect, and the random copolymers consisting of hydrophobic BA and zwitterionic A-Met(S+)-OH had a higher stabilizing effect at 4 °C than the poly(A-Met(S+)-OH) homopolymer.[26] In a previous work, the protein-stabilizing effect of the random copolymers was investigated during storage at 4 °C for HRP. In this study, the ability of the zwitterionic block copolymers to stabilize proteins was initially evaluated using HRP as an enzyme and 2,2′-azino-di-[3-ethylbenzthiazoline sulfonate] (ABTS) as a substrate in conditions resembling physiological conditions. It must be noted that the protein stability at a temperature higher than the room temperature is highly desirable for practical and industrial applications because of the feasibility to keep protein stability without a cooling system. Although HRP has numerous medical and industrial applications, its industrial applications have been limited by its instability under various conditions such as elevated temperatures and excess H2O2. The proteins were stored at 37 °C for several days in an aqueous solution containing 0.1% by weight of each sample as a protein stabilizer, which corresponds to an accelerated degradation test, whose effectiveness was evaluated by measuring the decrease in enzyme activity. The results of these protein stabilization tests are shown in Figure a,c. When no excipients were added, the activity of HRP was almost completely lost under the severe conditions. When the poly(A-Met(S+)-OH) homopolymer (P1) was used as a stabilizer, the activity was maintained at 90% under the conventional condition of 4 °C for 5 days,[26] whereas it decreased drastically to 5% under severe condition at 37° C for 5 days (Figure c). When the zwitterionic block copolymers with different A-Met(S+)-OH contents (BP1-5) were used as protein stabilizers, the protein-stabilizing effect was enhanced as the A-Met(S+)-OH content increased. In particular, the zwitterionic block copolymer with around 85% of A-Met(S+)-OH (BP2 and BP3) exhibited a significant effect on protein stability, with more than 60% of the activity remaining under severe conditions (37° C for 5 days), which is apparently higher than that obtained using BSA, a commonly used polymer-type protein stabilizer. In the case of the block copolymer (BP1) with the lowest hydrophobic segment (A-Met(S+)-OH/BA ratio = 96/4), the enzyme activity decreased drastically, suggesting that a specific amount of the hydrophobic component is required to achieve effective adsorption between the protein and polymer. In other words, a suitable hydrophilic/hydrophobic (zwitterionic A-Met(S+)-OH/nonionic BA) balance is crucial to obtain an effective protein-stabilizing effect of the sulfonium-containing zwitterionic block copolymers with characteristic amphiphilic property.

Figure 4

Activity of (a,c) HRP and (b,d) ALP after 5 days of storage at 37 °C with 0.1% by weight of (a,b) zwitterionic block copolymers (poly(A-Met(S+)-OH)-b-poly(BA)s) (BP1–BP5, Table ) and (c,d) random copolymers, poly(A-Met(S+)-OH-co-BA)s (CP1–4, Table S4, Supporting Information) (**p < 0.01 vs no additive, ##p < 0.01 vs BSA).

In order to clarify the effect of the comonomer sequence, the protein-stabilizing effect was compared between the block copolymers and random copolymers. The poly(A-Met(S+)-OH-co-BA)s, which were prepared through two-step RAFT copolymerization and postmodification by using iodomethane,[26] were employed for this purpose. In the case of random copolymers, poly(A-Met(S+)-OH-co-BA)s (CP1–4, Table S4, Supporting Information), the protein-stabilizing effect for HRP increases as the A-Met(S+)-OH content increases (Figure c), which is similar to the tendency observed for the block copolymers. This is the indication that the increase in the hydrophobic BA units resulted in a decrease in the enzyme activity, which may be due to the increase in the interaction between each protein and adsorption to hydrophobic surfaces. The comparison of the remaining activity between block and random copolymers (Figure a,c) indicates higher stabilization effect of the block copolymers, compared to that of the random copolymer, suggesting that the monomer sequence (location and distribution of hydrophobic BA units) is a crucial factor in determining the protein stability. Enzyme activity is generally decreased by contact between proteins because of interactions between the hydrophobic domains and adsorption to hydrophobic surfaces. Proteins in the polymer-type stabilizer solution interact with the hydrophobic site of the polymer, and therefore, the contact between proteins is suppressed and then stabilized.

To examine the effectiveness of these synthesized block copolymers as a protein-stabilizing agent, the stability of ALP was compared in the presence of the block copolymers, random copolymers, and homopolymers. Figure b,d exhibits the activity of ALP after 12 days of storage at 37 °C in the presence of 0.1% by weight of poly(A-Met(S+)-OH)-b-poly(BA)s (BP1–5) and poly(A-Met(S+)-OH-co-BA)s (CP1–4), respectively. In the presence of the zwitterionic block copolymers, relatively high ALP activity was maintained (>80%), independent of the A-Met(S+)-OH content of the block copolymers, under the same conditions. As shown in Figure , the zwitterionic A-Met(S+)-OH content-dependent stability of ALP is apparently distinct from that of HRP. There was a slight difference in the surface charges and molar mass between HRP (molar mass = 40 200 and zeta potential = −6.6 mV) and ALP (molar mass = 80 000 and zeta = −16.6 mV). In addition, the surface hydrophobicity of the enzymes should be related closely to the protein stabilization effect. Hence, a suitable hydrophilic/hydrophobic (zwitterionic A-Met(S+)-OH/nonionic BA) balance for protein stabilization is governed by the nature of the protein. In all cases, the block copolymers show higher stability compared to those of the random copolymers. Similar tendency of the higher stabilizing effect of block copolymers compared to random copolymers was observed in MPC-based zwitterionic copolymer systems.[36]

Figure 5

Activity of (a) HRP and (b) ALP against the composition of the zwitterionic A-Met(S+)-OH unit of the block and random copolymers.

Figure shows the postulated structures of a protein in the presence of block and random copolymers comprising of zwitterionic A-Met(S+)-OH and hydrophobic BA. When the zwitterionic block copolymer is used as a stabilizer, the surface of the protein–stabilizer complex is covered predominantly with a hydrophilic unit without significant defects. Hence, the contact between protein–stabilizer complexes is further suppressed, and a high stabilization effect is attained, if the poly(BA) chain is sufficiently long. In contrast, in the random copolymers, the hydrophilic A-Met(S+)-OH and the hydrophobic BA are randomly arranged. Therefore, the hydrophobic unit is also exposed on the surface of the protein–stabilizer complex, leading to increased interaction between the hydrophobic units. The increase in the BA content results in more contact between the protein–stabilizer complexes and decrease in enzyme activity. Another important factor for protein stabilization is the mobility of proteins inside the stabilizer, which is governed by the hydrophobic components around proteins.[36] Predominant presence of the poly(BA) segment at the protein interface can contribute to less movement of the protein, resulting in higher stability of the protein–block copolymer complex. In contrast, a certain amount of the zwitterionic A-Met(S+)-OH unit is located around the protein in random copolymers, which leads to the increase in the movement of the protein. Less contact between proteins due to the preferable presence of the zwitterionic poly(A-Met(S+)-OH) segment on the surface of the protein–stabilizer complex and less movement of the protein due to the predominant presence of the poly(BA) segment around the protein may contribute to higher protein-stabilizing effect in block copolymers compared to that in random copolymers. Note that the protein stabilization test was conducted at diluted solution (2.0 μg/mL), which is most probably lower than the critical micelle concentration. In terms of the effect of the comonomer composition, the increase in the hydrophobic BA content in the copolymers may lead to decrease in the mobility of the protein and decrease in the interaction between the protein–stabilizer complexes, which corresponds to the positive and negative effects on the protein stability, respectively. Such opposite effects of the incorporation of the hydrophobic BA unit on the protein stability are attributed to the presence of suitable hydrophilic/hydrophobic (zwitterionic A-Met(S+)-OH/nonionic BA) balance for protein stabilization depending on the nature of the protein. In addition to hydrophobic interaction, electrostatic interaction is another possible factor to affect protein stabilization. Nevertheless, the effect of the electrostatic interaction is not significant in the copolymers developed in this study because the zwitterionic A-Met(S+)-OH unit is electrically neutral.[26] Hence, the hydrophobic interaction is considered to be one of the major factors to contribute protein stabilization.

Figure 6

Schematic illustration of the postulated structures of a protein in the presence of block and random copolymers comprising of zwitterionic A-Met(S+)-OH and hydrophobic BA.

In the next stage, the protein stabilization effect of the zwitterionic block copolymer was evaluated in the presence of a sugar because saccharides (e.g., sucrose and trehalose) have been frequently used as low-molecular-weight protein-stabilizing compounds.[57] Here, the zwitterionic block copolymer showing the highest stabilization effect (BP3, A-Met(S+)-OH content = 85%) was selected as a polymeric stabilizer, and sucrose was used as a representative sugar-based stabilizer.

The protein stability test was evaluated using HRP, which was stored at 37 °C for 5 days in an aqueous solution containing 0.1% by weight of the zwitterionic block copolymer (BP3) in the presence and absence of sucrose (15% by weight). As shown in Figure , relatively low HRP was detected even in the presence of only sucrose, whereas a drastic increase in HRP activity was obtained by using a mixture of the zwitterionic block copolymers and sucrose. Structural stability via hydrogen bonding between the protein and sucrose may contribute to significant improvement in protein stability.[58−60] Sugars play a crucial role in the protection of cells and proteins by replacing the water for hydrogen bonding as well as the production of the glassy matrix. Another possible contributing factor is the exclusion volume of the sugars and the chemical nature of the protein surface.

Figure 7

Activity of HRP after 5 days of storage at 37 °C with 0.1% by weight of the zwitterionic block copolymer, poly(A-Met(S+)-OH)-b-poly(BA) (BP3), with 15% by weight of sucrose (**p < 0.01 vs no additive, ##p < 0.01 vs BP3).

Zwitterionic polymers are known to prevent protein aggregation by binding to hydrophobic sites on the protein surface, facilitating suitable polymer–protein interactions, which lead to the suppression of unfavorable protein–protein interactions and structural changes in the highly ordered structures.[9,61,62] In order to obtain further information on the interaction between the protein and the zwitterionic block copolymer, the hydrodynamic size, size distribution, and zeta potential of the protein in the presence and absence of the block copolymer were evaluated by DLS in the phosphate buffer (pH = 7.4) at 25 °C. To achieve sufficient peak intensity, DLS measurement was conducted in HRP solution (1 mg/mL) at a more concentrated condition used for the protein stabilization test (2.0 μg/mL). As shown in Figure S8 and Table S5 (Supporting Information), HRP solution exhibited two species, and a gradual change in the intensity-averaged hydrodynamic sizes (Dh) and their distributions was observed after the addition of the zwitterionic block copolymer (BP3, 0.1–1.0 mg/mL). Under the conditions, the zeta potentials of the polymer–stabilizer complexes were −3.52 and −3.63 mV at the HRP/polymer weight ratio = 1/0.1 and 1/1, respectively, which are between the values of pristine HRP (−6.64 mV) and block copolymer (−2.28 mV). The slight decrease in the zeta potential of HRP by the addition of the block copolymer indicates that the HRP is wrapped by the block copolymer. No significant difference in the zeta potential between the polymer–stabilizer complexes obtained at the HRP/polymer weight ratio = 1/0.1 and 1/1 suggests that the outermost surface of the polymer–stabilizer complexes is covered predominantly by the zwitterionic poly(A-Met(S+)-OH) segment, independent on the HRP/polymer weight ratio in the feed. In other words, a suitable selection of the concentration range of the zwitterionic block copolymer stabilizer is essential to achieve effective protein stabilization.

The zeta potential data of ALP in aqueous solution in the presence and absence of the amphiphilic block copolymer (BP3) at pH = 7.4 are shown in Figures S9 and S10 (Supporting Information). ALP was diluted with BP3 solution dissolved in ion exchange water (0.24 mg/mL), and then the zeta potential measurement was conducted. The zeta potential of pristine ALP was −16.6 mV, whereas the value of BP3 alone was almost zero (+0.5 mV). This is in good agreement with the general tendency of zwitterionic polymers, in which the zeta potential should be zero at the isoelectronic point. As BP3 was added to ALP, the zeta potential increased, as expected, suggesting that BP3 covers the surface of ALP via hydrophobic interaction. The zeta potential reached a plateau at a concentration of [BP3] = 0.016 mg/mL, where the zwitterionic block copolymer (BP3) may sufficiently covers the surface of the protein at the concentration of 1 mg/mL used in the protein stabilization test.

Transmission electron microscopy (TEM) was also employed to study the interaction between the HRP and zwitterionic block copolymer in phosphate-buffered saline (PBS). As shown in Figure a, spherical structures having a diameter of ca. 50–150 nm and their larger aggregated states composed of several spherical species are visible in the native HRP in PBS, which was stained with OsO4. Aggregated structures of HRP are also detected after the interaction with the amphiphilic block copolymer (BP3) at the HRP/polymer weight ratio = 1/0.1 (Figure b). Under the condition, crystalline-like structures were occasionally observed. In contrast, isolated spherical structures with more uniform size distribution are mainly seen when the sufficient amount of the block copolymer was employed (HRP/polymer weight ratio = 1/1, Figure c). Hence, TEM results suggest the preservation of the morphology and slight increase in the size of the native HRP in PBS by the addition of the zwitterionic block copolymer.

Figure 8

TEM photos of (a) 1 mg/mL of native HRP in PBS, (b) 1 mg/mL of HRP + 0.1 mg/mL of BP3 in PBS, and (c) 1 mg/mL of HRP + 1 mg/mL of BP3 in PBS.

The change in the conformation of the HRP–block copolymer relative to pristine HRP was evaluated by CD measurements in phosphate buffer (pH = 7.4) at 25 °C. HRP consists of 13 α-helices and 2 short antiparallel β-strands.[63] As shown in Figure , pristine HRP exhibited a strong negative peak at around 220 nm, which was attributed to the characteristic α-helix structure. No significant change was detected in the negative peak at around 220 nm after the addition of the zwitterionic block copolymer into the HRP solution, suggesting that the interaction with the block copolymer did not lead to any significant change in the enzyme helicity. The positive CD signal of HRP around 195 nm was shifted to around 205 nm, which may be due to the change in the hydrophilic–hydrophobic valence of the side chains by the interaction with the block copolymer. Nevertheless, these results consistently suggest that the zwitterionic block copolymer stabilizes an enzyme by wrapping with a slight change in the size, whereas most of the secondary and ordered structures of the enzyme are maintained.

Figure 9

Circular dichroism (CD) data of HRP, HRP with BP3, and BP3 in phosphate buffer (0.05 mg/mL).

Conclusions

In summary, we synthesized a series of sulfonium-containing zwitterionic block copolymers consisting of poly(A-Met(S+)-OH) and poly(BA) by RAFT polymerization and postmodification, in which the poly(BA) block can enrich the hydrophobic properties of these polymers and their compatibility with proteins, resulting in improved thermal stability. Initially, we found the conditions best suited toward controlled radical polymerization of the hydrophobic monomer (BA) from the hydrophilic poly(A-Met-OH) via the RAFT process in ethanol. The presynthesized poly(A-Met-OH)-b-poly(BA)s were successfully modified by reacting methyl iodide to afford targeted poly(A-Met(S+)-OH)-b-poly(BA)s. The RAFT polymerization of A-Met-OH using the hydrophobic poly(BA) macro-CTA and the subsequent postmodification were found to be another efficient approach for the controlled synthesis of poly(A-Met(S+)-OH)-b-poly(BA)s with different zwitterionic–nonionic (hydrophobic) compositions. The poly(A-Met(S+)-OH)-b-poly(BA) with 85% of the zwitterionic A-Met(S+)-OH content exhibited the highest stability for HRP at 37 °C for 5 days. The addition of sucrose into the zwitterionic block copolymer–protein solution led to a drastic increase in the HRP activity under the same conditions. In the presence of the zwitterionic block copolymers, the activity of ALP was maintained even after 12 days of storage at 37 °C. Tertiary sulfonium-containing zwitterionic copolymers consisting of A-Met(S+)-OH and BA had excellent properties to enhance the stability of proteins by manipulating the sequence of the monomer units (block and random) and hydrophilic/hydrophobic (zwitterionic A-Met(S+)-OH/nonionic BA) balance originating from the comonomer composition.

Experimental Section

Materials

l-Methionine (>99%), acryloyl chloride (>98%), iodomethane (>99.5%), BA (>99%), trimethylsilyldiazomethane (0.6 M in hexane solution), and p-nitrophenyl phosphate (pNPP, ready-to-use solution) were purchased from Tokyo Chemical Industry and were used without purification. N-Acryloyl-l-methionine (A-Met-OH) was prepared by the reaction of acryloyl chloride with l-methionine according to a method reported previously.[26,64,65] The monomer was purified by column chromatography on silica with ethyl acetate/methanol (1:1, v/v) as the eluent.

The radical initiator AIBN (>98%) was purchased from FUJIFILM Wako Pure Chemical Corporation and was used without purification. Benzyl 1-pyrrolecarbodithioate (97%), as a dithiocarbamate-type CTA, and BSA (>98%) were purchased from Aldrich and were used directly. Ethanol (>99.5%) and MTBE (>99%) were purchased from Kanto Chemical Corporation.

For protein stabilization tests, HRP (>100 units/mg, molar mass = 40 200) and ALP (40–80 units/mg, molar mass = 80 000) were purchased from FUJIFILM Wako Pure Chemical Corporation. The ABTS microwell peroxidase substrate (one-component system) was purchased from SeraCare Life Sciences, Inc. All other materials were purchased and used without purification.

Synthesis of Amphiphilic Block Copolymers, Poly(A-Met(S+)-OH)-b-poly(BA)

The synthesis of the block copolymer employing poly(A-Met-OH) as the macro-CTA was performed in accordance with Scheme a. A-Met-OH (4.28 g, 20.0 mmol), CTA (93.2 mg, 0.40 mmol), AIBN (32.8 mg, 0.20 mmol), and DMF (20 mL) were placed in a dry round-bottom flask equipped with a magnetic stirring bar, and the solution was degassed by N2 gas bubbling for 0.5 h. The mixture was stirred at 65 °C for 3 h, and the resulting polymer was precipitated into MTBE. Conversion of double bonds was 77% as detected by 1H NMR. The yellowish solid obtained (2.74 g, 64%) was subjected to 1H NMR (CD3OD, 400 MHz), yielding the following values: δ 4.3–4.6 (1H, NH–CH–COOH), 2.5–2.8 (2H, CH2–CH2–S), 2.0–2.6 (5H, CH2–CH2–S, S–CH3), 1.6–2.6 (polymer backbone). 13C NMR (CD3OD, 100 MHz) values were as follows: δ 175.6 (COOH), 175.3 (C(=O)NH), 52.9 (CH–COOH), 43.2 (polymer backbone), 37.0 (polymer backbone), 31.7 (CH–CH2–CH2–S), 27.3 (CH2–S–CH3), 15.7 (S–CH3).

SEC was conducted on the methylated sample poly(A-Met-OMe), which was obtained by treating the carboxylic acid groups of the resulting poly(A-Met-OH) with trimethylsilyldiazomethane. The methylation of the poly(A-Met-OH)s was carried out according to a previously reported method.[66,67] The solvents were then removed by evaporation, and the methylated samples were measured by SEC, without any purification, using DMF Mn = 8200 and Mw/Mn = 1.23 in the methylated form, which corresponds to Mn = 7700 in the carboxylic acid-containing poly(A-Met-OH). Note that trimethylsilyldiazomethane must be accompanied by appropriate safety precautions, and the evaporation process should be carried out with proper ventilation. Quantitative esterification (>99%) was confirmed by comparing the integration of the methyl resonance at 3.5–3.8 ppm (−COOCH3) with the intensity of the methylene resonance at 2.4–2.7 ppm (CH2–CH2–S). 1H NMR (400 MHz, CD3OD) values were as follows: δ 4.4–4.6 (1H: NH–CH–COOH), 3.5–3.8 (3H: -COOCH3), 2.4–2.7 (2H, CH2–CH2–S), 2.0–2.6 (5H, CH2–CH2–S, S–CH3). The 1H NMR spectra of poly(A-Met-OH) and poly(A-Met-OMe) are shown in Figure S1 (see the Supporting Information).

For the synthesis of the block copolymer (BP4), the dithiocarbamate-terminated poly(A-Met-OH) (Mn,NMR = 12 000 and Mw/Mn = 1.23, 236 mg, 20.0 μmol), BA (298 mg, 2.30 mmol), AIBN (1.60 mg, 10.0 μmol), and ethanol (2.3.0 mL) were placed in a dry glass tube equipped with a magnetic stirring bar, and the solution was degassed by N2 gas bubbling for 0.5 h. The mixture was stirred at 65 °C for 20 h. The BA conversion was determined by the integration of the monomer C=C–H resonance at ≈5.9 ppm compared to the sum of the (C=O)–OCH2– peak intensities for the polymer and monomer at ≈4.8 ppm. The BA conversion was found to be 78%. The resulting block copolymer, poly(A-Met-OH)-b-poly(BA), was purified by reprecipitation from ethanol solution into a large excess of MTBE. The product was dried under vacuum at room temperature to yield a yellow solid product (468 mg, 77%). The comonomer composition of the anionic–hydrophobic block copolymer (A-Met-OH/BA molar ratio = 31/69) was determined using 1H NMR spectroscopy by comparison of the peaks corresponding to the two comonomers (with the peak at 2.5–2.8 ppm attributed to the S–CH2– proton and the peak at 1.0 ppm corresponding to the −CH2–CH3 protons). 1H NMR (CD3OD, 400 MHz) values are as follows: δ 4.4–4.8 (1H, NH–CH–COOH), 3.8–4.3 (2H, (C=O)–OCH2), 2.5–2.8 (2H, CH2–CH2–S), 2.2–2.5 (2H, CH2–CH2–S) 2.0–2.2 (3H, S–CH3), 1.6 (2H, (C=O)–O–CH2–CH2), 1.4 (2H, CH2–CH2–CH3), 1.0 (3H, CH2–CH3), 1.2–2.6 (polymer backbone). The solubility of the anionic–hydrophobic block copolymer in various solvents is summarized in Table S1 (see the Supporting Information).

Similar to the homopolymer, the methylation of the poly(A-Met-OH) block in the resulting poly(A-Met-OH)-b-poly(BA) was conducted by treating the carboxylic acid groups with trimethylsilyldiazomethane for SEC measurement. An excess of the methylation agent (0.6 M in hexane solution, 100 μL, 60 μmol) was added to 300 μL of tetrahydrofuran/MeOH (2:1 v/v) solution of the block copolymer (10 mg, which corresponds to 41.5 μmol of A-Met-OH unit), and the solution was stirred for another 4 h at room temperature. After removing the solvents by evaporation, the methylated sample was measured by SEC without any purification. The methylated poly(A-Met-OMe)-b-poly(BA) displayed an Mn (as determined by SEC) of 11 000 and a polydispersity index of 1.41.

The zwitterionic block copolymer, poly(A-Met(S+)-OH)-b-poly(BA), was obtained by treating the sulfide group of the A-Met-OH component in the carboxylic acid-containing amphiphilic block copolymer, poly(A-Met-OH)-b-poly(BA), with iodomethane. For the block copolymer ([BA]/[macro-CTA] = 115/1, BP4), poly(A-Met-OH)-b-poly(BA) (200 mg) was dissolved in ethanol (3 mL), and 0.5 mL of iodomethane ([iodomethane]/[A-Met-OH] = 9/1) was added. After stirring at room temperature for 20 h in an N2 environment, the poly(A-Met(S+)-OH)-b-poly(BA) was further dialyzed against methanol for 6 h, followed by dialysis against H2O for 1 day (MWCO: 1000 Da). The copolymer solution was lyophilized to yield a yellowish powdery product (205 mg, 95%). 1H NMR (CD3OD + 20% D2O, 400 MHz) values were as follows: δ 4.3–4.7 (1H, NH–CH–COOH), 3.8–4.3 (2H, (C=O)–CH2), 3.4–3.8 (2H, CH2–CH2–S(CH3)2), 3.0–3.2 (6H, CH2–S(CH3)2), 1.2–2.8 (CH2–CH2–S, polymer backbone), 1.6 (2H, (C=O)–CH2–CH2), 1.2 (2H, CH2–CH2–CH3), 0.9 (3H, CH2–CH3). The 1H NMR spectra of the carboxylic acid-containing and zwitterionic block copolymers, poly(A-Met-OH)-b-poly(BA) and poly(A-Met(S+)-OH)-b-poly(BA), are shown in Figure . The solubility of the zwitterionic block copolymers in various solvents is summarized in Table S2 (see the Supporting Information).

The zwitterionic–hydrophobic block copolymer (BP5) was also synthesized by RAFT polymerization of A-Met-OH from poly(BA) macro-CTA, followed by similar modification of the methyl thioether moiety (Scheme b). The procedure is detailed in the Supporting Information. The 1H NMR spectra of the poly(BA) macro-CTA, poly(BA)-b-poly(A-Met-OH), and poly(BA)-b-poly(A-Met(S+)-OH) are shown in Figure S3 (Supporting Information).

Protein Stabilization Study (Accelerated Degradation Test)

Protein Stabilization Test with HRP

HRP solution (2.0 μg/mL) was prepared in PBS pH 7.4. The samples were dissolved in PBS pH 7.4 at a concentration of 0.1% by weight. Aliquots (50 μL) of the HRP solution were added to the sample solutions (1 mL) at 4 °C. Sample solutions containing HRP and an additive-free control sample were stored at 37 °C for several days. All samples were assessed immediately after preparation. ABTS was used as the substrate, and the reaction of HRP was carried out at ambient temperature for 30 min. Then, the reaction was stopped using 1 wt % sodium dodecyl sulfate solution. Absorbance was measured at 410 nm to monitor activity, and the assay was repeated thrice. All p-values were calculated using the Tukey–Kramer test.

Protein Stabilization Test with ALP

An ALP solution (1.2 mg/mL) was prepared in Tris-HCl buffer saline pH 9.0 containing 1.0 mM MgCl2. The samples were dissolved in Tris-HCl buffer saline pH 8.0 containing 1.0 mM MgCl2 at a concentration of 0.1% by weight, and 4 μL of aliquots were added to the sample solutions (1 mL) at 4 °C. Sample solutions containing ALP and an additive-free control sample were stored at 37 °C for several days. All samples were assessed immediately after preparation. pNPP was used as the substrate, and the reaction of HRP was carried out at ambient temperature for 30 min. Then, the reaction was stopped using 2 M sodium hydroxide solution. Absorbance was measured at 405 nm to monitor activity, with the assay being repeated thrice. All p-values were calculated using the Tukey–Kramer test.

Instrumentation

The 1H (400 MHz) and 13C (100 MHz) spectra were recorded using a JEOL JNM-AL400. Mn and Mw/Mn were estimated using SEC at 40 °C using a Tosoh HPLC HLC-8020 system equipped with refractive index and ultraviolet detectors. For poly(A-Met-OMe) and poly(A-Met-OMe)-b-poly(BA), the column set was as follows: two consecutive columns [Polymer Laboratories (PLgel) (bead size, exclusion-limited molar mass): mixed-D (5 μm, 4 × 105) × 2] and a guard column [Polymer Laboratories (PLgel) guard column]. The system was operated at a flow rate of 0.6 mL/min, and DMF containing 1 g/L of LiBr was used as the eluent. Poly(methyl methacrylate) standards were employed for calibration. Zeta potential data were obtained on a Zetasizer Nano-ZS (Malvern) in H2O. Block copolymer micelles were formed by gradual addition of deionized water (5 mL) to a solution of the polymer (5 mg) in 50 μL of MeOH + 20% H2O, affording a mixed solvent (water/MeOH = 124/1 vol %). Prior to the light scattering measurements, the polymer solutions were filtered using Millipore Teflon filters with a pore size of 0.8 μm into a dust-free cylindrical cuvette. Zeta potential data were obtained on a Zetasizer Nano-ZS (Malvern) in H2O. The pH was adjusted using sodium hydroxide and hydrogen chloride. The CD was measured using a JASCO J-720 spectropolarimeter. Measurements in the far-ultraviolet region (180–250 nm) were performed on sample solutions in pH 7.0, 10 mM sodium phosphate buffer employing a 1.0 mm path-length cell at 25 °C. The CD spectra were recorded under the same conditions running two scans for each sample, corrected by subtracting the appropriate blank runs.

TEM observation was performed on a JEOL TEM-2100F field emission electron microscope at an accelerating voltage of 200 kV. The sample for TEM observation was prepared by mounting a drop of the mixed sample dissolved in PBS with an adjusted concentration on carbon-coated Cu grids. After the solvent was evaporated at room temperature for 3 h, aliquots of an aqueous solution of OsO4 (2 wt %) were dropped onto the dried grit, and the sample was stained for 3 min. Then, the sample was washed with hexane and dried for overnight. The glass-transition points (Tg) of the copolymers were measured by DSC (DSC 6220, Seiko Instruments Inc.). After annealing at 120 °C, the samples were cooled to −80 °C at a cooling rate of 10 °C/min and subsequently scanned at a heating rate of 10 °C/min from −80 to 120 °C.