Galactose-Functional RAFT Agent for Biorelevant Polymeric Materials

Abstract

In this work, we report a novel chain transfer agent (CTA) derived from d-galactose and evaluate its potential as a reversible addition–fragmentation chain-transfer (RAFT) agent. Using this galactose-based CTA, Gal-RAFT, both hydrophilic and hydrophobic monomers were successfully polymerized, yielding amphiphilic polymers, demonstrating its broad applicability. Additionally, we investigated the optimal reaction conditions, including initiator type, concentration, solvent, and temperature. The resulting galactose-containing amphiphilic polymer, Gal-pDMAEMA-b-pBzMA, self-assembled into polymeric nanoparticles (<150 nm, D _h) in aqueous media. These amphiphilic polymeric nanoparticles were characterized and evaluated for biorelevance, with lectin-binding studies demonstrating surface biofunctionalization and selective molecular recognition. Results of this study aid in establishing a versatile synthetic platform for end-functional macromolecules that enables next-generation design of bioactive and functionally adaptive polymeric materials.

Introduction

Reversible addition–fragmentation chain-transfer polymerization (RAFT) is considered one of the most versatile and finely tunable strategies for synthesizing well-defined polymers across diverse architectures and functional domains. In RAFT polymerization, the resulting polymers typically exhibit low dispersity, primarily because the chain transfer agent (CTA) establishes a rapid equilibrium between active and dormant polymer chains. The CTA is comprised of Z and R groups, which are crucial for controlling polymer molecular weight and enabling diverse architectures (Figure ). The R group of the CTA plays a pivotal role in governing polymerization and fragmentation rates, functioning as the reinitiating species and thereby significantly influencing both polymerization kinetics and the capacity for block copolymer formation. If the R group is overly stable, fragmentation is slowed, resulting in polymerization retardation; conversely, if it is too unstable, reinitiation is inefficient, compromising the overall control of the polymerization process. The Z group of the CTA plays a critical role in the polymerization process by modulating the stability of the radical intermediate and determining the range of monomers that can be polymerized efficiently. Electron-withdrawing Z groups, such as aryl or carbonyl-containing moieties, preferentially facilitate the polymerization of more activated monomers like methacrylates, whereas electron-donating Z groups (e.g., alkylthio) are suitable for less activated monomers, such as vinyl esters. Among RAFT reagents, dithioesters and trithiocarbonates are particularly effective at controlling the polymerization of more activated monomers (MAMs). In these agents, the Z group is connected via a carbon atom in dithioesters and via a sulfur atom in trithiocarbonates, a structural feature that significantly influences their polymerization behavior.

1.

Representation of the Z and R groups on the Gal-RAFT chain transfer agent (CTA).

Recent advances demonstrate that incorporating specialized moieties into the Z group enables direct control over polymerization kinetics and facilitates the synthesis of polymers with tailored functionality. This provides access to materials with precisely engineered properties, including responsiveness, biocompatibility, and site-specificity. Foster et al. reported dithiocarbamate CTAs bearing norbornene-containing Z groups that facilitate controlled polymerization of specific monomer classes. These findings underscore the pivotal role of Z-group structure in modulating RAFT efficiency and enabling dual polymerization strategies. −

In this study, we present a novel RAFT reagent by introducing d-galactose directly as the Z group in RAFT polymerization (Figure ). The scope of this technique has been extended to functional polymers, including glycopolymers, for several biomedical applications. Traditionally, saccharide units have been introduced into polymers solely as monomers , and hardly as components of RAFT reagents. To the best of our knowledge, this research represents the first report of incorporating a d-galactose unit directly into the RAFT reagent.

Glycosylated polymers have long been recognized for their critical roles in biological processes and their potential in the development of innovative therapeutic agents for biomedical applications. , Moreover, glycopolymers can be engineered as biosensors via carbohydrate–lectin interactions, which underlie numerous biological processes, and their ability to mimic natural carbohydrate–protein interactions further highlights their potential in biomedical research and therapeutics. − Specifically, galactose is uniquely suited for polymer functionalization due to its high-affinity, receptor-specific recognition by the asialoglycoprotein receptor (ASGPR) on hepatocytes and by galectins on cancer cells, enabling targeted cellular uptake rather than passive metabolic interactions. , Unlike other monosaccharides, galactose combines receptor selectivity, biocompatibility, and low immunogenicity, and when displayed multivalently on polymer chains, exhibits amplified binding through the cluster glycoside effect. Ladmiral et al. reported the use of galactose-decorated monomers to generate biocompatible nano-objects (spheres, worms, vesicles) via RAFT polymerization, targeting receptors to deliver cargo intracellularly, thereby highlighting carbohydrate–protein specificity for biomedical delivery. In addition to biomedical uses, galactose-based gelators have been explored for phase-selective solvent separation and cationic dye removal, demonstrating their wider applicability in remediation and related areas. Qui et al. demonstrated the use of galactose-based polymers as a powerful tool for fabricating assemblies with tunable morphologies. The galactose-functionalized methacrylate polymer synthesized via RAFT polymerization formed a highly ordered “glyco-inside”, thereby broadening prospects for targeted delivery and biosensing applications. Consistent with this rationale, our previous work also demonstrated that galactose can be used to directly govern nanoparticle properties, underscoring its critical role as a selective and effective targeting ligand for biofunctional polymer systems.

Adding biologically important molecules, such as galactose, as the Z group in RAFT reagents is a particularly innovative approach that expands the role of saccharides beyond their traditional function as monomers and creates inherently bioactive CTAs. This design enables controlled polymerization while providing a direct platform for glyco-functionalization, opening new opportunities in drug delivery, biomaterials, and glycopolymer research. Herein, to incorporate the galactose moiety as the Z group in RAFT reagents, a galactose-based RAFT reagent was synthesized and evaluated for the polymerization of both hydrophilic and hydrophobic monomers. Using this reagent, hydrophilic, hydrophobic, and amphiphilic polymers were synthesized by optimizing conditions to achieve the desired molecular weights and controlled dispersity, suitable for biomedical applications. These polymers were characterized using various techniques, confirming that the optimized conditions enabled precise control of polymerization with this innovative RAFT system. The results demonstrate the utility of this approach for forming saccharide-coated nanoparticles, significantly broadening the applicability of RAFT polymerization and transforming it from a versatile technique into a platform for designing next-generation functional materials for nanomedicine.

Experimental and Methods

d-Galactose, lithium aluminum hydride (LAH) in 2 M THF, 2-bromo-2-methyl propionic acid, benzyl methacrylate, 2-(dimethylamino)­ethyl methacrylate (DMAEMA), AIBN (2,2′-Azobis­(2-methylpropionitrile)), carbon disulfide, and K₃PO₄ were procured from Sigma, 2-(hydroxymethyl)-2-methylpropanoic acid, 2-(hydroxyethyl)­methacrylate, para toluene sulfonic chloride (pTSCl), V50 (2,2′-Azobis­(2-methylpropionamidine) dihydrochloride), sulfuric acid were supplied by Thermo Fisher Scientific. Potassium thioacetate was obtained from TCI. All the solvents used were anhydrous. All chemicals were used as received, unless otherwise specified. Additional detailed descriptions of synthesis and product characterization, nanofabrication results, and biorelevant assessment are provided in the Supporting Information.

General Equipment

Nuclear Magnetic Resonance Spectroscopy

¹H NMR spectroscopy was performed using a 400 MHz Bruker AVANCE III spectrometer. Chain transfer agent and monomer spectra were acquired with 32–64 coadded scans and a delay time of 5 s. Polymer spectra were acquired using 64 coadded scans and a delay time of 2 s. All spectra were obtained using the appropriate deuterated solvents (CDCl₃, DMSO-d ₆, or MeOH-d ₄) and were processed and analyzed using MNova software.

Gel Permeation Chromatography/Size-Exclusion Chromatography

Protected glycopolymer weight- and number-average molecular weights were determined using an Agilent Technologies 1260 Infinity multidetector suite (refractive index detector only) and an Agilent Technologies 1260 ISO pump fitted with a 1260 ALS autosampler. The gel permeation chromatography/size-exclusion chromatography (GPC/SEC) eluent was HPLC-grade DMF containing 0.01% LiBr and was filtered prior to use. GPC/SEC was equipped with a 300 × 8 mm PSS GRAM analytical column and 50 × 8 mm guard column. HPLC-grade DMF was used as the eluent at a flow rate of 0.5 mL/min, operating at 55 °C with a sample concentration of 5 mg/mL and an injection volume of 100 μL. The instrument was calibrated with linear poly­(methyl methacrylate) standards (800–40k). All samples were passed through 0.2 μm nylon filters prior to GPC/SEC measurements.

Preparation and Characterization of Nanoparticles

Polymer systems were formed into nanoparticles via a previously reported nanoprecipitation method. 1 mg of the polymer was dissolved in THF (200 μL), and the resulting solution was added dropwise to a vial containing Milli-Q water (2 mL) while sonicating. THF was allowed to evaporate under constant nitrogen flow. Nanoformulations were allowed to equilibrate for 12 h before testing.

Particle size and zeta potential (ζ-potential) measurements were carried out on a Zetasizer Nano ZS (Malvern Instrument) using a He–Ne laser (633 nm) detector angle of 173° at 25 °C. Concentration was maintained at 0.5 mg/mL for all systems, and all measurements were performed in triplicate to ensure consistency.

Transmission Electron Microscopy

Nanoparticle solutions were deposited on pretreated EM grids (Lacey Carbon, EMS, LC325-CU-25; grids were glow discharged in the EasiGlow system for 1 min at 20 mA to ensure the surface was hydrophilic). After plating, the grids were flash-frozen using a manual cryo-plunger. The frozen grids were screened using Thermo Scientific Glacios Cryo-TEM. Grids with suitable ice thickness were transferred to a Thermo Scientific Titan Krios for cryo-ET data collection. The Titan Krios was equipped with an AMETEK Gatan K3 direct detector and AMETEK Gatan energy filter. Cryo-ET data were collected at 300 kV at a magnification of 57,000×.

DLS-Based Lectin Binding Assessment

To assess binding of different nanoparticles with peanut agglutinin (PNA) lectin, nanoparticles (1 mg/mL) were incubated with different PNA concentrations (1, 0.75, 0.5, and 0.25 mg/mL) at a 1:1 (v/v) ratio in buffer (50 mM NaCl, 10 nM HEPES, 0.1 mM CaCl₂, 0.01 mM MnCl₂ at pH 7.5) for 60 min at room temperature. Dynamic light scattering (DLS) measurements were performed before and after incubation to monitor changes in hydrodynamic diameter and PDI. Protected and galactose-free nanoparticles were used as negative controls.

Fluorescence Plate-Based Lectin Binding Assay

Noncovalent immobilization of nanoparticles onto a black polystyrene 96-well plate was performed by incubating 1 mg/mL nanoparticle suspension at different volumes (80, 60, 40, 20, and 10 μL), with the total volume made up to 100 μL with buffer, overnight at 4 °C. The next day, wells were washed three times with buffer and incubated with FITC-labeled PNA (1 mg/mL) in buffer for 60 min at room temperature. Unbound lectin was removed by washing three times with buffer. Buffer (100 μL) was then added to each well and fluorescence emission spectra (513–600 nm) were measured using a Thermo Scientific Varioskan LUX multimode plate reader at an excitation wavelength of 495 nm. Uncoated wells incubated with FITC-PNA and unmodified NP-coated wells incubated with FITC-PNA were used as controls to assess nonspecific adsorption of lectin onto the plate surface or NPs. Fluorescence intensities were normalized to the corresponding NP-only and uncoated well measurements to account for any autofluorescence and nonspecific binding. A 10-point moving average was applied to smooth the final spectra.

Cytotoxicity Assay

HEK-293 cells were grown under standard conditions (37 °C and 5% CO₂) in DMEM med. Cytotoxicity associated with particles was determined by CCK-8 assay. Cells in 96-well plate were exposed to a range of particle concentrations for 24 h followed by 2 h incubation with CCK-8 reagent (ApexBio). Cell viability was assessed by measuring the absorbance at 450 nm.

Kinetics Experiments

Kinetic studies were performed to monitor monomer conversion during RAFT polymerization. Polymerizations were conducted under the standard conditions as described below. At predetermined time intervals, small aliquots (0.20 mL) of the reaction mixture were withdrawn using a degassed syringe under a nitrogen atmosphere. Each aliquot was rapidly quenched by cooling and exposed to air. The quenched samples were diluted in CDCl₃ and analyzed by ¹H NMR spectroscopy. Monomer conversion was determined by integrating the polymer benzylic proton signals at 4.8 ppm relative to the monomer anomeric proton signal at 5.4 ppm. The spectral peaks of the aliquots were analyzed to monitor the evolution of molecular weight (M _n as a function of conversion) and the degree of polymerization (DP). Plots of ln­([M]₀/[M]) vs time were generated to evaluate pseudo-first-order kinetics, while M _n and D̵ vs monomer conversion were plotted to assess molecular weight evolution and control over dispersity.

Results and Discussion

Synthesis of Gal-RAFT Reagent

The RAFT reagent governs the control, versatility, and fidelity of the polymerization process. The structure of the RAFT reagent typically contains a thiocarbonylthio moiety (Z–C­(S)–S–R). The Z and R groups play a crucial role in regulating chain growth by maintaining a dynamic equilibrium between active and dormant chains. Fine-tuning these two groups enables the RAFT reagent to produce polymers with predictable molecular weights, narrow molecular weight distributions, and complex architectures. Here, we developed a novel galactose-based RAFT reagent in which a protected galactose unit serves as the Z group and an acid group with a spacer as the R group (Figure ). To the best of our knowledge, this is the first report of a multifunctional d-galactose-based CTA (termed, Gal-RAFT).

Considering its ability to interact with specific receptors, ,, we introduced the galactose unit into the RAFT reagent, making it a valuable building block for creating site-specific therapeutic agents. The synthesis of the Gal-RAFT reagent involved five steps and afforded the target in moderate yield (Scheme ). The synthesis of thiolated galactose was based on an established protocol where the secondary hydroxyls of d-galactose were protected using acetone in the presence of a catalytic amount of sulfuric acid to achieve (1) (yield 82%). The RAFT reagent was then synthesized by coupling with the 2,2-dimethyl bromo propionic acid, followed by tosylation (2) (yield 83%) and thiolation (yield 81%) (3) using tosyl chloride and thioacetate, respectively. Deacetylation of the thioacetate afforded intermediate (4) (yield 56%) (Scheme ). The compound (4) was further treated with 2-bromomethyl propionic acid to obtain the desired galactose-based RAFT reagent (5) (yield 88%). Product 5 was confirmed by ¹H and ¹³C NMR spectroscopy. Proton signals for the protected galactose moiety were observed at 5.44, 4.55, 4.21, and 3.94 ppm, while the peaks at 3.57 and 3.33 ppm corresponded to the methyl protons adjacent to the thiocarbonylthio group CH₂–S–CS–S–. Resonance signals in the region between 1.24 and 1.36 ppm confirmed the presence of the isopropylidene protecting group (CH₃)₂C– on the galactose unit. Additionally, a doublet at 1.65 ppm was assigned to the methyl group (CH₃)₂–S–CS–S– located near the acid functionality (Supporting Information). Collectively, these results confirmed the successful synthesis of the novel Gal-RAFT reagent.

1. Stepwise Synthesis Scheme of the Gal-RAFT Reagent.

Polymerization Using Gal-RAFT Reagent

A protected Gal-RAFT reagent was employed in this study to introduce a structurally defined and functionalized end group without interfering with the RAFT polymerization mechanism. The presence of protective groups (e.g., acetyl or benzyl) renders the galactose moiety hydrophobic, enhancing the solubility of the RAFT reagent in organic solvents such as DMF or methanol. Additionally, the bulky saccharide structure can impart subtle steric effects that may improve chain-end stability and reduce termination events. Notably, the galactose unit remains chemically latent during polymerization and can be deprotected postsynthetically to yield a bioactive saccharide-functional polymer chain end.

Influence of Initiator on Polymerization

The potential of the novel Gal-RAFT reagent was evaluated using benzyl methacrylate (BzMA) as a model monomer with V-50 (hydrophilic) and AIBN (hydrophobic) initiators, targeting polymers with a molecular weight of 10k. BzMA was selected as the model monomer because it is the most widely used monomer in RAFT polymerization and enables the formation of polymers with controlled molecular weight, specific morphologies, and properties. Importantly, BzMA polymers exhibit favorable mechanical and self-assembling properties, thereby preventing the kinetic trapping of self-assembled molecules.

V-50 was examined at multiple concentrations (0.1, 0.05, 0.02, and 0.01 M) and in different solvent systems (DMF, MeOH, and MeOH/DMF mixtures at 50 and 70 °C) at a fixed [M]/[CTA] ratio of 54:1 (Supporting Information Table S1). Across all conditions, the polymerizations exhibited poor molecular weight control, broad dispersities, or negligible conversion. These observations are attributed to the limited solubility of V-50 in organic media, the decreased polarity associated with hydrophobic BzMA, and an elevated radical flux , at higher temperatures, leading to uncontrolled chain initiation. , Reactions conducted in anhydrous methanol proceeded more slowly and likewise failed to improve control, likely due to hydrogen-bonding effects that reduce methacrylate propagation, particularly for hydrophobic monomers that rely on optimal solvation.

Based on the results from V-50, we employed an alternative initiator, AIBN. AIBN is a common, hydrophobic initiator and typically works well in organic solvents (Scheme and Table ). We maintained the initiator concentration at 0.1 M, the solution concentration at 25% w/v, and the reaction temperature at 50 °C. The evolution of molecular weight was studied at multiple time points. After 5 h, the dispersity (D̵) of the polymer formed was 2.00, and the M _n (18,507) and M _w (37,092) values remained higher, which did not align with the theoretical molecular weight (10,000). The reaction was further incubated overnight, and the M _n and M _w values were slightly lower than those of the 5 h samples (14,626 and 34,430, respectively). However, the dispersity (2.35) was higher than that of the polymerization reaction after 5 h.

2. Synthesis of Gal-pBzMA Polymer Using Gal-RAFT Reagent General Scheme; Initiator Concentration, Temperature, and Solvent Conditions Detailed in Table .

1. Influence of Reaction Conditions on Polymerization of BzMA Using Gal-RAFT Reagent with AIBN .

						GPC
trial	initiator equiv. (I)	solvent	solution conc. (w/v) %	temperature (°C)	time (h)	M n	M w	D̵
1	0.1	DMF	25	50	5	18,507	37,092	2
2	0.1	DMF	25	50	ON	14,626	34,430	2.35
3	0.1	dioxane	50	70	2	19,000	22,277	1.17
4	0.1	dioxane	50	70	4	13,897	18,391	1.32
5	0.15	dioxane	50	70	3	10,000	16,317	1.55
6	0.15	dioxane	50	70	ON	19,000	22,536	1.14
7	0.1	dioxane	50	70	1	16,498	22,695	1.37
8	0.1	dioxane	50	70	2	14,713	23,365	1.58
9	0.1	dioxane	50	70	3	14,165	23,560	1.66
10	0.1	dioxane	50	70	4	14,301	24,179	1.69
11	0.05	dioxane	50	70	3	14,154	20,202	1.43
12	0.05	dioxane	50	70	4	14,487	20,478	1.42
13	0.05	dioxane	50	70	5	15,311	21,304	1.39
14	0.075	dioxane	50	70	1	12,055	17,509	1.45
15	0.075	dioxane	50	70	2	12,338	17,928	1.45
16	0.075	dioxane	50	70	3	12,400	17,944	1.45
17	0.075	dioxane	50	70	4	13,556	19,205	1.42
18	0.075	dioxane	50	70	5	14,149	20,150	1.42

^aTheoretical molecular weight, Mn_th: 10,000; conditions: M: I: CTA = 1: I*: 54, solvent used is anhydrous, ON: overnight samples (12–13 h) (where I* = 0.15/0.1/0.075/0.05).

Interestingly, when the RAFT polymerization of BzMA in anhydrous DMF at 70 °C was allowed to proceed overnight, a slight decrease in molecular weight (M _n) was observed compared to earlier time points. This phenomenon is consistent with RAFT end-group degradation and bimolecular termination events, which are known to occur under prolonged radical exposure and elevated temperatures.

As reported by Moad et al., extended heating in polar solvents can compromise the thiocarbonylthio end group of RAFT reagents, leading to chain fragmentation or loss of living character. , Perrier also highlighted that when monomer conversion is high, continued radical generation can cause termination between active chains, producing dead polymers and shifting the molecular weight distribution. Additionally, radical-induced backbone scission reactions may occur during prolonged thermal exposure, especially in methacrylate systems in polar environments.

We next changed the solvent to dioxane and increased the solvent concentration to 50% w/v, while maintaining the initiator concentration, the monomer-to-initiator ratio, and the temperature as in the previous reaction. The change in solvent and its concentration resulted in polymerization within 2 h. The molecular weight, as determined by SEC/GPC (M _n = 19,000, M _w = 22,277), was higher than theoretical, but the D̵ (1.17) was lower (Table ). The reaction was allowed to proceed for 4 h, after which the molecular weight decreased (M _n = 13,897, M _w = 18,391), accompanied by an increase in the D̵ (1.32). This decrease in molecular weight can be attributed to several factors mentioned above, including chain fragmentation and Gal-RAFT reagent degradation.

Notably, when the initiator concentration was increased to 0.15 M under identical conditions at different time points, the M _n value obtained after 3 h via SEC/GPC agreed with the theoretical molecular weight. However, the polymer obtained exhibited a higher D̵ of 1.55. For the overnight sample, the molecular weight increased as expected. However, the dispersity was significantly improved (1.14), indicating that a higher radical flux under these conditions may have overcome oxygen inhibition or maintained a sufficient population of active chains, thereby delaying degradation and supporting continued propagation. Over time, as more chains propagate and dormant chains re-enter the RAFT equilibrium, the system undergoes chain-length equalization, resulting in a narrower molecular weight distribution. These observations are consistent with recent findings by Taylor et al., who demonstrated that moderately high initiator concentrations in RAFT polymerization promote oxygen tolerance and lead to improved polymer control with time.

The effect of varying AIBN concentration on the RAFT polymerization of BzMA was evaluated under constant conditions: 70 °C, 50% w/v solvent concentration, and galactose-based RAFT reagent. At a lower initiator concentration (0.05 M), polymerization was only observed after 3 h, indicating delayed initiation, likely due to insufficient radical flux. , In contrast, increasing the initiator concentration to 0.075 or 0.1 M led to rapid polymerization, with initiation occurring within the first hour. The molecular weight increased with time at lower initiator concentrations (0.05 and 0.075 M), but showed poor control, with dispersity values ranging from 1.39 to 1.45. At a higher initiator concentration (0.1 M), the molecular weight decreased over time, and the dispersity broadened significantly, increasing from 1.37 to 1.69 after 4 h. These results confirm that AIBN concentration critically influences not only the rate of polymerization initiation but also molecular weight buildup and control, with higher concentration leading to increased radical termination events and reduced control over the RAFT process.

Based on these findings, we proceeded with an initiator concentration of 0.075 M and aimed to achieve a target molecular weight (10k) by reducing the solvent concentration to 30% (w/v). The reaction was maintained for ∼6 h. The ¹H NMR spectrum displayed a broad peak corresponding to the benzyl proton signal (−CH₂–O−) at 4.80 ppm and the aromatic proton signals (−CH−) merged at 7.21 ppm, confirming the polymerization (Figures , S12 and S13). A doublet at 5.40 ppm (−CH–O–; anomeric hydrogen) indicated the presence of the Gal-RAFT reagent and remained stable during the reaction, as no additional peak was observed.

2.

General ¹H NMR spectrum of hydrophobic polymer Gal-pBzMA (Mn_th = 30k, n _NMR = 176), solvent CDCl₃; additional spectra provided in Supporting Information (Figures S12 and S13).

We expanded our synthesis toward targeted molecular weights of 20k with [M]: [CTA]: [I] = 111:1:0.075 and 30k with [M]: [CTA]: [I] = 167:1:0.075 respectively under the same optimized conditions used for the 10k polymer (i.e., 0.075 M AIBN concentration and 30% w/v dioxane). In both cases, we achieved polymers with molecular weights in close agreement with theoretical values and narrow dispersity indices (Table ), indicating good control over polymerization. We successfully achieved targeted molecular weights of 10, 20, and 30k (Mn_SEC = 10,931, 22,310, and 29,074) with well-dispersed dispersity indices (Table ). Spectroscopic characterization of all polymers (Figures , S12 and S13) confirmed that the protected galactose-functionalized RAFT reagent was suitable for the controlled polymerization of hydrophobic monomers, such as benzyl methacrylate. The improved control is attributed to an optimized balance between radical flux and monomer concentration, as well as the consistent chain-transfer behavior of the Gal-RAFT reagent. Its efficient chain-transfer behavior and compatibility with the reaction medium support the synthesis of well-defined polymers across a range of target molecular weights. Although the galactose moiety was protected, it provides a chemically versatile handle for postpolymerization modifications, making the system suitable for the design of functional glycopolymers in future studies.

2. Differing Molecular Weights for Gal-pBzMA Were Achieved Using Optimized Reaction Conditions, Including Monomer Conversion, NMR-Determined Molecular Weight, and SEC/GPC-Measured M _n and Dispersities .

	Mnth	MnNMR	MnSEC	D̵	[M]o:[CTA]o	conversion %	DP (NMR)
1.	10,000	11,792	10,931	1.05	54:1	98	67
2.	20,000	20,768	22,310	1.2	111:1	99	118
3.	30,000	30,976	29,074	1.08	167:1	99	176

^aMonomer conversions determined by ¹H NMR spectroscopy; determined by SEC/GPC, refractive index detector; additional data provided in Supporting Information (Figures S12–S15).

Kinetics Study Using BzMA Monomer

The kinetics of RAFT polymerization of BzMA involve several key steps: initiation, pre-equilibrium, reinitiation, main equilibrium, propagation, and termination. , Factors such as solvent composition, RAFT reagent concentration, and temperature influence each step. Considering this, we assessed the kinetics for the polymerization of BzMA in anhydrous dioxane at 70 °C with [I]: [M]: [CTA] = 0.075:54:1, and a targeted molecular weight of 10k. The polymerization of BzMA using the Gal-RAFT followed the conventional/classic RAFT polymerization kinetics. The rate of polymerization was influenced by initiator concentration, monomer concentration, the propagation rate constant (k _p), and the structures of the leaving and stabilizing groups of the CTA. , Figure displays the kinetic profile of the Gal-pBzMA homopolymer with 98% conversion within 5–6 h. The kinetic plot displays a generally increasing trend but was not perfectly linear over the full conversion range (Figure S14). It is noted that the RAFT agent designed herein is structurally distinct, and the observed deviations may reflect inherent features of this system. , Although deviations from ideal pseudo-first-order behavior were observed, the polymerization exhibited controlled molecular weight growth and acceptable dispersity values, supporting the ability of the Gal-RAFT reagent to mediate polymerization under the investigated conditions. ,

3.

¹H NMR spectra of the reaction of Gal-RAFT reagent with BzMA to form pBzMA (Mn_th = 10k). (a) Kinetics using ¹H NMR as monomer conversion versus time profile of Gal-pBzMA polymerization (b). (Additional plot of degree of polymerization vs conversion provided in the Supporting Information (Figure S14)).

Influence of Monomer on Polymerization

After applying the galactose-based RAFT reagent for the polymerization of BzMA, a hydrophobic monomer, we optimized the performance of the galactose-based RAFT reagent for the polymerization of hydrophilic monomers (Scheme ). DMAEMA was selected as the model hydrophilic monomer because its tertiary amine imparts pH- and temperature-responsive behavior, and it is highly soluble in aqueous/organic media, and readily undergoes controlled radical polymerization, affording stimuli-responsive platforms and functional polymer architectures.

3. Synthesis of Hydrophilic Gal-pDMAEMA Polymer Using Gal-RAFT Reagent, 0.075 equiv. AIBN, 30% w/v Solvent Dry Dioxane 70 °C.

Similar reaction conditions were maintained for the hydrophilic monomer. AIBN was used as the initiator, anhydrous dioxane as the solvent, with a 30% w/v solution concentration with [M]: [CTA]: [I] = 124:1:0.075. The reaction was maintained at 70 °C for 5 h, with a theoretical molecular weight of 20k. NMR confirmed the polymerization of DMAEMA using the Gal-RAFT reagent, with a broad peak at 4.04 ppm corresponding to the –CH₂–OCO group, a peak at 2.57 ppm corresponding to the –CH₂–N– linkage, and a peak at 2.28 ppm corresponding to the (CH₃)₂–N– group. A peak at 3.85 ppm (−CH₂–S–), and a peak at 1.86 ppm (CH₃–C­(CO₂R)–S–) corresponded to the carbon thiol linkage from the RAFT reagent to the inserted monomer (Figure ). The molecular weight was calculated by using NMR and GPC/SEC affording Mn (NMR), M _n and M _w values of 21,528, 18,263, and 22,060 with a dispersity of 1.21, respectively (Figure S15). Although the GPC/SEC displayed a unimodal peak in the NMR spectrum, it showed an additional doublet at 5.49 ppm corresponding to the anomeric proton of galactose, indicating partial degradation of the RAFT reagent. As a result, some polymer chains may be terminated prematurely, accounting for the slight reduction in molecular weight observed by GPC/SEC relative to the theoretical value. The degradation is likely due to the nature of DMAEMA, which may promote side reactions such as aminolysis or base-catalyzed hydrolysis of the RAFT end-group under thermal conditions. These findings align with previous reports in which monomers bearing basic functional groups have been shown to compromise RAFT end-group integrity, resulting in chain termination and lower-than-expected molecular weights. Hafliger et al. reported that prolonged heating can degrade RAFT end groups, leading to a measurable decrease in molecular weight. Similarly, Stace et al. observed that RAFT polymers lose their thiocarbonyl end-groups at temperatures below those required for backbone degradation, notably affecting molecular weight control. Overall, these observations highlight the influence of monomer functionality on the performance of the Gal-RAFT reagent.

4.

¹H NMR of hydrophilic polymer Gal-pDMAEMA (Mn_th = 20k, n _NMR = 138), solvent CDCl₃.

Block Copolymer Synthesis

After evaluating RAFT polymerization of the Gal-RAFT reagent with individual monomers, we synthesized an amphiphilic block copolymer of benzyl methacrylate (BzMA, hydrophobic) and DMAEMA (hydrophilic) (Scheme ). The 10k BzMA macro-CTA was chain-extended with DMAEMA (target 10k) to yield a 20k amphiphilic block copolymer. The polymerization was carried out in dioxane (30% w/v), 0.075 equiv. initiator, at 70 °C under previously optimized conditions.

4. Synthesis of Block Copolymer Gal-pDMAEMA-b-pBzMA, and AIBN 0.075 equiv., 30% w/v Dry Dioxane, 70 °C for 5.5 h (Mn_th = 20k, n = 56, m = 64).

NMR confirmed the polymerization (Figure ), with the values indicating the presence of the amphiphilic block. The benzyl proton (−CH₂–O−) signal appears at 4.88 ppm, and the aromatic proton (−CH−) signal at 7.27 ppm, confirming the hydrophobic unit. A broad peak at 4.06 ppm corresponding to the –CH₂–OCO group, a peak at 2.56 ppm corresponding to the –CH₂–N– linkage, and a peak at 2.28 ppm corresponding to the (CH₃)₂–N– group, confirmed the hydrophilic unit. A peak at 5.50 ppm representing the anomeric hydrogen indicated the presence of the CTA. The degree of polymerization (DP) determined by NMR was 67 hydrophobic units and 59 hydrophilic units, yielding a molecular weight of 21,055, which is slightly higher than the theoretical value (20k). In comparison, the molecular weights obtained by SEC/GPC were consistent with the theoretical value, with M _n = 20,453 and M _w = 24,544. The SEC/GPC peak was unimodal, indicating that both units had similar molecular weights (Figure S15).

5.

¹H NMR of block copolymer Gal-pDMAEMA-b-pBzMA (Mn_th = 20k, n _NMR = 67, m = 59), solvent CDCl₃.

Deprotection and Self-Assembly of Gal-pDMAEMA-b-pBzMA Block Copolymers

With the amphiphilic polymer in hand, nanoparticles were formed via polymer self-assembly. Based on block architecture, micelles are expected, with pBzMA forming the core and pDMAEMA constituting the corona. Particle size was analyzed in aqueous media using DLS, which provides the hydrodynamic diameter (D _h). For this study, the polymer nanoparticles were characterized in both Milli-Q water and physiological buffers, as size, charge, and stability depend strongly on solution conditions. , Measurements in water reflect intrinsic properties, while buffer studies are expected to reveal behavior under biologically relevant conditions. In Milli-Q water, protected Gal-pDMAEMA-b-pBzMA has a particle size with a Z-average diameter of 190.0 ± 1.53 nm, with 0.11 PDI and a zeta potential of 27.7 ± 0.23 mV (Figure S16). In buffer, particle size was smaller with a Z-average diameter of 62.6 ± 0.26 nm, 0.07 dispersity, and a zeta potential of 21.9 ± 0.52 mV (Figure S18). In parallel, the block copolymer Gal-pDMAEMA-b-pBzMA was treated with 8% (v/v) TFA in chloroform for 4 h to deprotect the isopropylidene groups present on the galactose moiety. A noticeable difference was observed in the nanoparticle solutions: the protected Gal-pDMAEMA-b-pBzMA afforded a cloudy suspension, whereas the deprotected polymer yielded a transparent nanoparticle solution (Supporting Information). Those nanoparticles formed in Milli-Q water resulted in a Z-average diameter of 102.1 ± 0.44 nm, with PDI of 0.096 and a z-potential of 25.9 ± 0.70 mV (Figure S17). Again, in buffer, particle size was notably smaller with a Z-average diameter of 49.4 ± 0.26 nm, 0.11 dispersity, and a zeta potential of 24.6 ± 0.88 mV (Figure S19). The smaller nanoparticle sizes observed in buffer likely result from reduced hydration and slight contraction of the polymer corona. Dissolved salts weaken polymer–water interactions, causing the hydrated shell to shrink and lowering the apparent size measured by DLS.

Cryogenic transmission electron microscopy (CryoTEM, Figures , S20 and S21) confirmed spherical morphologies for both systems but revealed substantially smaller NP sizes, with protected nanoparticles averaging 40.6 nm and deprotected nanoparticles averaging 29.9 nm (Figure ). While the size trends remain consistent between DLS and CryoTEM, it can be speculated that the size reduction observed by CyroTEM is a common yet fundamental difference between the two methods may also reflect genuine structural reorganization. The size of the nanoparticles decreased upon deprotection, due to the removal of the isopropylidene groups. These protecting groups can lead to aggregation of the polymer chains and/or adoption of a more extended conformation, thereby increasing the hydrodynamic volume. Whereas deprotection produced a more compact, soluble polymer, resulting in a smaller D _h in DLS measurements. As such, deprotection likely increases polymer polarity and hydrogen-bonding capacity while reducing steric bulk, affording tighter chain packing. A similar result was obtained by Wu et al., who found that deprotecting the acetate group on the sugars led to shrinkage of the nanoparticles. The zeta potential values after deprotection of the galactose unit decreased slightly, most likely due to water molecules forming a hydrophilic, neutral layer that may shield the charges, thereby reducing the measured zeta potential.

6.

Cryogenic transmission electron microscopy (cryoTEM) images of protected (a) and deprotected (b) Gal-pDMAEMA-b-pBzMA nanoparticles; scale bar = 100 nm; additional micrographs included in Supporting Information.

To directly confirm the presence of galactose on the nanoparticle surface, we performed a comparative analysis of protected, deprotected, and galactose-free control polymers (PDMAEMA-b-PBzMA) using lectin-binding assays. For this study, the galactose-free control polymer, PDMAEMA-b-PBzMA, and its nanoparticles were included to account for potential cationic interactions arising from the PDMAEMA block (Supporting Information). DLS studies were conducted across a range of lectin (PNA, peanut agglutinin or Arachis hypogaea) concentrations (0.25–1 mg/mL) at constant nanoparticle concentration to probe carbohydrate–lectin interactions in solution (Figures S24–S26). The deprotected Gal-pDMAEMA-b-pBzMA nanoparticles exhibited a pronounced, concentration-dependent increase in hydrodynamic diameter from 49.4 to 86.6 nm (Figure S24) along with a substantial broadening of the polydispersity index (PDI: 0.11 to 0.56), consistent with specific lectin–saccharide recognition driving aggregation. This behavior is attributed to the tetrameric nature of PNA, which enables multivalent bridging between galactose-presenting nanoparticles, generating heterogeneous NP–PNA assemblies with varying stoichiometries and resulting in broadened size distributions. In contrast, protected Gal-pDMAEMA-b-pBzMA nanoparticles showed no measurable change in hydrodynamic diameter (∼55–56 nm) or PDI (<0.10) across all lectin concentrations, indicating that surface modifications effectively shield galactose moieties and prevent lectin access (Figure S25). Similarly, galactose-free nanoparticles derived from PDMAEMA-b-PBzMA exhibited minimal or inconsistent size changes, with only a modest increase at the highest lectin concentration (1 mg/mL), likely due to nonspecific electrostatic interactions rather than specific binding (Figure S26). Importantly, no significant changes were observed at lower PNA concentrations (≤0.75 mg/mL), confirming the absence of carbohydrate recognition.

Complementary fluorescence measurements further supported these findings, collectively providing strong evidence that galactose is selectively presented on the surface of deprotected nanoparticles and is accessible for specific biomolecular interactions. Here, a fluorescence-based lectin-binding assay was developed to evaluate PNA binding to surface-immobilized nanoparticles. The fluorescence spectra were recorded for all three nanoparticles across a range of concentrations (10–80 μL), with all spectra normalized to nanoparticles-only and lectin-only controls to account for polymer autofluorescence and non-specific binding of the FITC-labeled PNA, respectively (Figure ).

7.

Fluorescence-based lectin binding assay evaluating the lectin binding to surface immobilized nanoparticles: deprotected Gal-pDMAEMA-b-pBzMA (a); protected Gal-pDMAEMA-b-pBzMA (b); control PDMAEMA-b-PBzMA (c); and normalized comparison at their highest tested concentration (d). Supporting DLS results provided in Supporting Information, Figures S23–S26.

Deprotected Gal-pDMAEMA-b-pBzMA exhibited a concentration-dependent FITC emission across the 513–600 nm range (Figure a). The presence of fluorescence signal confirms retention of the galactose FITC conjugate even after extensive washing steps. Protected Gal-pDMAEMA-b-pBzMA similarly exhibited concentration-dependent FITC emission, with spectral profiles comparable to those of the unprotected samples. The difference in spectral intensity profile from deprotected nanoparticles reflects differences in the accessibility of the PNA conjugated FITC fluorophore due to surface modification (Figure b). Control PDMAEMA-b-PBzMA nanoparticles also showed lower fluorescence emission across all concentrations tested compared to deprotected nanoparticles, consistent with the absence of galactose conjugation in this formulation (Figure c). The low level of fluorescence in pDMAEMA-b-pBzMA nanoparticles likely reflects nonspecific binding of nanoparticles and lectin similar to DLS studies. Normalized comparison of all three nanoparticles at their highest tested concentration (80 μL, Figure d) confirmed that deprotected Gal-pDMAEMA-b-pBzMA exhibited the highest fluorescence. This data collectively support specific galactose-mediated PNA recognition in deprotected Gal-pDMAEMA-b-pBzMA.

To further assess the applicability of the deprotected Gal-pDMAEMA-b-pBzMA nanoparticles, cytotoxicity was evaluated using HEK-293 cells, a rapidly proliferating, low-maintenance cell line routinely employed for preliminary biocompatibility screening. Cells exposed to Gal-pDMAEMA-b-pBzMA nanoparticles showed no measurable cytotoxicity at concentrations up to 32 μg mL^–1 (Figure S27), indicating appropriate in vitro tolerance at doses relevant for biomedical applications. At higher concentrations, approximately 30% cytotoxicity was observed, suggesting a concentration-dependent response rather than acute toxicity. Such findings are unusual for a biomolecular conjugate, suggesting that additional modifications may be needed to reduce high-dose cytotoxicity while preserving desirable nanoscale properties. , Nonetheless, compared with commercially established platforms, these results are comparable to early stage safety profiles reported before surface optimization or payload incorporation. , Collectively, this data supports the potential of the Gal-RAFT CTA and its ability to form amphiphilic polymers suitable for therapeutic applications.

Conclusions

Here, we present a novel galactose-based CTA and demonstrate its utility in RAFT polymerization. Incorporation of galactose directly into the Z-group enables controlled synthesis of hydrophobic, hydrophilic, and amphiphilic methacrylate polymers with predictable molecular weights, low dispersities, and successful block copolymer formation. Reaction conditions employing AIBN (0.075 M) and 30% w/v dioxane were found suitable for the polymerization of the BzMA reaction using the Gal-RAFT reagent. Noting that the type of monomer and temperature can affect the molecular weight built up during RAFT polymerization, we juxtaposed the polymerization behavior of Gal-RAFT with that of hydrophobic and hydrophilic methacrylate monomers. Polymerization of BzMA yielded molecular weights close to theoretical values and relatively narrow dispersities under the optimized conditions, consistent with monomer solubility and stable RAFT equilibria in organic media. In the case of DMAEMA polymerization, slightly broader dispersity and molecular weights lower than expected were observed. ¹H NMR analysis indicated partial degradation of the RAFT end group, likely due to interactions between the tertiary amine functionality of DMAEMA and the thiocarbonylthio moiety under the reaction conditions. Formation of an amphiphilic polymer (Gal-pDMAEMA-b-pBzMA) underwent spontaneous self-assembly in aqueous media, while postpolymerization deprotection of the galactose unit offers a means to tune nanoparticle size and interfacial properties. Light scattering (DLS) analysis revealed nanoparticle sizes within the therapeutic window, while lectin-binding assays confirmed that galactose is present on the surface of deprotected nanoparticles and is accessible for specific biomolecular interactions. By coupling precise polymerization control with inherent biofunctionality at the chain end, Gal-RAFT offers a versatile platform for the design of glyco-functional materials and establishes a general strategy for integrating biologically relevant motifs into precision polymer synthesis.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acspolymersau.6c00036.

• Detailed descriptions of synthesis and product characterization, nanofabrication results, and biorelevant assessment are provided accordingly (PDF)

⊥.

A.K. and S.M. share co-first authorship and contributed equally to experimental optimization, data collection, and data interpretation. A.K. initiated the experiments and outlined them, while S.M. optimized those experiments and troubleshot and finalized data collection. All authors have given approval to the final version of the manuscript.

The Ohio State University and The University of Mississippi provided the funding for this work.

The authors declare no competing financial interest.