Introduction

Extracellular polysaccharides (EPS) are a group of organic polymeric compounds released into the surrounding environment by microorganisms [1]. Recently, EPS or exopolysaccharides have been considered interesting bio-based materials, which show great potential for application in wastewater treatment systems, owing to their ability to absorb and biodegrade certain substances [2]. They are also widely used in the cosmetics, food, and pharmaceutical industries because of their good safety profile and physical and rheological properties [3, 4]. In addition, polysaccharide-based amphiphilic copolymers play a crucial role in gene and drug delivery systems because of their ability to spontaneously self-aggregate to form nanoscale structures such as micelles [5, 6]. These micelles enhance drug solubility and stability by effectively encapsulating the hydrophobic compounds in their cores. Compared with synthetic copolymers which are potentially cytotoxic and are closely linked to the depletion of petroleum resources, polysaccharide-based copolymers possess better biocompatibility, biodegradability, and abundant hydrophilic hydroxyl groups for chemical modification or functionalization [7].

However, the presence of polysaccharides in biological tissues, such as cellulose in plant cell walls, presents challenges during purification. Acquiring pure polysaccharides requires intricate and multistep manipulations, substantial energy input, and the use of organic solvents [8]. Furthermore, these elaborate production and purification procedures result in the biggest drawback of polysaccharides, namely, high polydispersity in the composition between different batches. To overcome these limitations and produce new, synthetic polysaccharides, genetic engineering has been employed to develop enzymes capable of synthesizing polysaccharides with high purity and low polydispersity in an environmentally friendly manner [9,10,11].

α-1,3-Glucan is an α-1,3-linked homopolymer of glucose and is mainly found in fungal cell walls and in biofilms produced by Streptococcus species [12, 13]. Recent studies have reported that water-insoluble α-1,3-glucan can be synthesized in high yield from sucrose under mild aqueous conditions; its production process is environmentally friendly and inexpensive because α-1,3-glucan can be easily purified by precipitation without the use of organic solvents. This synthesis was achieved through in vitro enzymatic polymerization using glucosyltransferase I (GTF-I) or glucosyltransferase J (GTF-J), which are enzymes genetically engineered from Streptococcus species [9, 14]. Thus far, α-1,3-glucan is a promising new low-cost bio-based material, and its application in bioplastic or hydrogel fields has been widely discussed. Our previous report has introduced biocompatible poly(L-lactide) into α-1,3-glucan backbone via homogeneous ring-opening polymerization (ROP) in ionic liquid, 1-butyl-3-methylimidazolium chloride (BmimCl) [15]. Notably, the amphiphilic α-1,3-glucan derivatives were able to form spherical nanomicelles ranging from 57 to 125 nm by self-assembly in an aqueous solution and showed high dilution stability with low CMC values (35 to 6 mg L−1). Furthermore, these nanomicelles were suitable for application as hydrophobic drug delivery vehicles [15].

ε-Caprolactone (ε-CL), a hydrophobic aliphatic polyester, possesses nontoxicity, low immunogenicity, and excellent biodegradability and biocompatibility [16,17,18,19]. Although it has been widely applied as the hydrophobic component in amphiphilic copolymers for micelle formation, a grafted brush copolymer based on ε-caprolactone and α-1,3-glucan has not yet been explored. The combination of α-1,3-glucan with the biodegradable poly(ε-caprolactone) (PCL) holds promise in producing amphiphilic graft copolymers and nanomicelles suitable for drug delivery, thereby broadening the scope of α-1,3-glucan applications in the biomedical field. In this study, a novel amphiphilic α-1,3-glucan derivative (Glucan-g-PCL) was synthesized via the graft modification of α-1,3-glucan with ε-CL. The self-assembly of the Glucan-g-PCL copolymer was performed in water to obtain a micelle-like nanostructure with PCL inner core and α-1,3-glucan outer layer. Quercetin (Qu), a plant-derived polyphenol commonly found in fruits and vegetables, has limited bioavailability and stability owing to its hydrophobic nature [20]. We investigated the encapsulating and solubilizing capacities of the Glucan-g-PCL nanostructure for Qu, as a poorly water-soluble model compound. Furthermore, the stability, antioxidant activity, and in vitro release behaviors of Qu-loaded Glucan-g-PCL nanostructure were also determined. Enzymatic polymerization-derived α-1,3-glucan based biodegradable amphiphilic copolymers exhibit great potential for application as drug nanocarriers.

Experimental procedures

Materials

1-Butyl-3-methylimidazolium chloride (BmimCl) with > 98.0% purity, 4-dimethylaminopyridine (DMAP) with > 99.0% purity, ε-caprolactone (ε-CL) with > 99.0% purity, and quercetin hydrate with > 96.0% purity were purchased from Tokyo Chemical Industry Co. Ltd. (Japan). All other reagents were of analytical grade, obtained from FUJIFILM Wako Pure Chemical Corporation (Japan), and were directly used as received without further purification. α-1,3-Glucan was produced by enzymatic polymerization according to the method described in the literature [15].

Synthesis of Glucan-g-PCL copolymer

Glucan-g-PCL was synthesized according to a previously reported procedure with a slight modification [17]. Briefly, dried α-1,3-glucan (0.4 g) and BmimCl (10.0 g) were successively added to a 100-mL two-neck flask, and the mixture was kept at 100 °C for 1 h with vigorous magnetic stirring. Then, the monomer of ε-CL (for α-1,3-glucan: ε-CL weight ratio of 1:10) and 0.2 wt% DMAP were slowly added, and the temperature was increased to 110 °C. The mixture was kept under a nitrogen atmosphere and further stirred for 24 h. The mixture was cooled to room temperature. The mixture was added to 300 mL of ethanol, followed by filtration with a 0.45-μm glass filter (Millex®, Merck, Germany) to isolate the resulting Glucan-g-PCL copolymer. Then, the isolated product was washed three times with ethanol and stirred in 30 mL of dichloromethane at room temperature for 72 h to dissolve the homo-PCL by-product. After filtration through a glass filter, the purified Glucan-g-PCL copolymer was washed several times with more dichloromethane in order to confirm the absence of any free homo-polymer and dried in a desiccator for 48 h.

Glucan-g-PCL micelle-like nanostructure preparation

Glucan-g-PCL micelle-like nanostructure was prepared in an aqueous solution using a dialysis method. Briefly, 10 mg of the Glucan-g-PCL copolymer was dissolved in 1 mL of DMSO, and the solution was added into 10 mL of deionized water under vigorous stirring. The solution was transferred to a dialysis bag (MWCO = 3500) and dialyzed against water for 48 h and the deionized water was refreshed six times in order to remove DMSO. After dialysis, the polymeric nanostructure solutions were passed through a 0.45-μm membrane filter and adjusted to a definite concentration for subsequent analysis.

Qu loading and in vitro Qu release

The Qu-loaded Glucan-g-PCL (Qu/Glucan-g-PCL) nanostructure solutions were prepared using a method similar to the emulsion solvent evaporation method [21, 22]. Briefly, 10 mL of Glucan-g-PCL nanostructure solution (1 mg mL−1) was prepared as described above. Subsequently, a predetermined amount of Qu dissolved in the THF solution was slowly dropped into the Glucan-g-PCL nanostructure solutions under vigorous stirring. To encapsulate Qu into the core of the nanostructure, the mixture was sonicated for 30 min on ice using a probe-type sonicator. Then, the remaining THF solvent was completely removed from the mixture solutions at 37 °C by rotary evaporation and purging with nitrogen gas. Finally, the solutions were adjusted to a certain concentration by adding deionized water and then passed through a 0.45-μm membrane filter to remove unloaded Qu. The concentration of loaded Qu was calculated using a standard curve. The encapsulation efficiency (EE, %) and loading content (LC, %) of Qu in the Glucan-g-PCL nanostructure were measured at 373 nm using a spectrophotometer (U-2900, Hitachi, Japan) according to the following equations:

$$\mathrm{EE}(\%)=\left(\mathrm{weight}\;\mathrm{of}\;\mathrm{loaded}\;\mathrm{Qu}\;/\;\mathrm{weight}\;\mathrm{of}\;\mathrm{feed}\;\mathrm{Qu}\right)\;\times\;100\;\%$$
$$\mathrm{LC}(\%)=\left(\mathrm{weight}\;\mathrm{of}\;\mathrm{loaded}\;\mathrm{Qu}\;/\;\mathrm{weight}\;\mathrm{of}\;\mathrm{polymers}\right)\;\times\;100\;\%$$

In vitro Qu release profiles of the Qu/Glucan-g-PCL nanostructure were examined in phosphate-buffered saline (PBS) solution of pH 7.4 at 37 °C. Briefly, 4 mL of fresh Qu/Glucan-g-PCL nanostructure solution was placed into a dialysis bag (MWCO = 3500) and dialyzed against 30 mL PBS solution (pH 7.4) containing 1% Tween-80 at 37 °C with gentle shaking at 100 rpm. At predetermined time intervals, 3 mL of the PBS solution was removed in a timely manner to measure the absorbance intensity at 373 nm using a spectrophotometer (Hitachi U-2900), and 3 mL of fresh buffer solution was added. The cumulative amount of released Qu was calculated using the following equation.

$$\mathrm{Cumulative\; release}\left(\mathrm{\%}\right)=(\mathrm{weight\; of\; released\; }/\mathrm{\;weight\; in\; micelles})\times 100\%$$

Measurements and characterization

FT-IR and 1H NMR measurement

The Fourier transform infrared (FT-IR) spectra of the α-1,3-glucan and Glucan-g-PCL copolymer were recorded using a FT-IR spectrometer (FT/IR-680 Plus, Jasco, Japan). Place a small amount of sample (approximately 3–7 mg) into the sample cell, and the measurement was performed in the range of 4000 to 500 cm−1 at room temperature. Meanwhile, in order to verify the loading of the Qu and the interaction between the loaded Qu and Glucan-g-PCL nanostructure, under the conditions of Qu equilibrium amount (both samples contain the same amount of Qu), the FT-IR spectra of physical mixture of Qu/Glucan-g-PCL copolymer and freeze-dried Qu/Glucan-g-PCL nanostructure were also recorded in the range of 4000 to 500 cm−1 at room temperature. 1H nuclear magnetic resonance (NMR) spectra measurements were recorded using a spectrometer (JNM-ECS 400, Jeol, Japan) at 25 °C. Samples were prepared for NMR analysis by dissolving 10 mg of each sample in 1 mL of dimethyl sulfoxide-d6 (DMSO-d6) with 3% LiCl.

X-ray diffraction measurement

The crystal structure of each sample was characterized using X-ray diffraction (XRD) spectroscopy on a D8 DISCOVER (Bruker AXS, USA), in which the high-intensity monochromatic nickel-filtered Cu Kα radiation was generated at 40 kV and 40 mA. The samples were scanned at a speed of 2°/min in a 2θ range of 5–45° with a step size of 0.04° at room temperature.

Thermogravimetric analysis

The thermal stability of each sample was evaluated using thermogravimetric analysis (TGA) (Thermo Plus Evo Rigaku, Japan) in the dynamic mode under nitrogen flow. Each sample (8 mg) was heated in an aluminum crucible from room temperature to 500 °C at a heating rate of 10 °C/min.

Size distribution and morphological analysis

The particle size and distribution of Glucan-g-PCL nanostructure and Qu/Glucan-g-PCL nanostructure were determined by dynamic light scattering (DLS; DLS-8000, Otsuka Electronics Co., Ltd.) instrument at a 90° scattering angle with a 632.8 nm He–Ne laser using backscattering detection. All samples were tested at least five times, and the average statistical values were recorded. The morphological features of Glucan-g-PCL nanostructure and Qu/Glucan-g-PCL nanostructure were investigated by transmission electron microscopy (TEM, JEM-2100 Plus, Japan) operating at an accelerating voltage of 150 kV. Samples were prepared as follows: one drop of the nanostructure solutions with a concentration of 0.5 mg mL−1 on a 300-mesh carbon-coated copper grid, stained with 2 w/w% phosphotungstic acid solution, and dried at room temperature before testing.

Critical aggregation concentration measurement

The critical aggregation concentration (CAC) values of the Glucan-g-PCL copolymer were determined using pyrene as a hydrophobic probe via fluorescence spectroscopy (RF-5300PC, Shimadzu, Japan) [23]. Glucan-g-PCL nanostructure solutions with a series of concentrations ranging from 0.001 to 1 mg mL−1 were prepared. The pyrene/acetone solution (6 × 10−5 M, 50 μL) was added to 5 mL of each solution, and then solvent acetone was completely evaporated under nitrogen. The final concentration of pyrene in the nanostructure solutions was maintained at 6 × 10−7 M. Then mixed solutions were sonicated using an ultrasonic cleaning machine for 30 min at room temperature to equilibrate the pyrene. The spectra were recorded from 360 to 550 nm at an excitation wavelength of 339 nm. The peak height intensity ratio (I1/I3) of the first peak (I1 at 373 nm) to the third peak (I3 at 384 nm) was plotted against the logarithm of polymer concentration.

Antioxidant activity

The antioxidant activity of Qu encapsulated in Glucan-g-PCL nanostructure was studied using the 2,2-diphenyl-1-picrylhydrazyl (DPPH) method [24]. A determined amount of untreated Qu powder, α-1,3-glucan, Glucan-g-PCL nanostructure, and Qu/Glucan-g-PCL nanostructure were added to 30 mL conical tubes, and water was added water to keep the total volume consistent. Samples for DPPH radical scavenging were prepared following filtration through a 0.45-μm membrane filter. Then, a freshly prepared 1 mL DPPH/ethanol solution (80 μg mL−1) was mixed with each sample or deionized water (blank sample). The mixtures were incubated for 30 min in the dark at room temperature. The absorbance of the solution was measured at 517 nm by using a spectrophotometer (Hitachi U-2900). DPPH scavenging was evaluated using the following equation:

$$\mathrm{DPPH\; scavenging }\;\left(\mathrm{\%}\right)=({{\text{Absorbance}}}_{{\text{blank}}}-{{\text{Absorbance}}}_{{\text{sample}}}/{{\text{Absorbance}}}_{{\text{blank}}})\times 100\mathrm{\%}$$

Stability of Qu/Glucan-g-PCL nanostructure

The prepared blank nanostructure solution and Qu/Glucan-g-PCL nanostructure solution were stored at 4 °C for 4, 8, 12, 16, 20, and 25 days, respectively. After the storage time was reached, the stability (particle size and polydispersity index (PDI)) of the samples was investigated by DLS, and the concentration of Qu entrapped in nanostructure during storage was measured at 373 nm using a spectrophotometer (Hitachi U-2900). The degradation rate was calculated using the following equation:

$$\mathrm{Retention\; rate }\;\left(\mathrm{\%}\right)=({C}_{0}-{C}_{t}/{C}_{o})\times 100\mathrm{\%}$$

where C0 and Ct are the initial and final concentrations, respectively, after storage of Qu in Glucan-g-PCL nanostructure [21].

Results and discussion

Synthesis of α-1,3-glucan and Glucan-g-PCL copolymer

The production scheme of α-1,3-glucan and the amphiphilic α-1,3-Glucan-graft-PCL copolymer are shown in Fig. 1. α-1,3-Glucan as a starting material was produced via in vitro enzymatic polymerization with recombinant GTF-I from Streptococcus mutans ATCC 700610 following the same method reported in our previous paper [15]. Because of the simplified extraction process of α-1,3-glucan and because the need for harmful organic solvents was eliminated, this method of enzymatic polymerization for α-1,3-glucan is considered to be environmentally friendly [9, 25].

Fig. 1
figure 1

Schematic depiction for the synthesis of α-1,3-glucan (a) and α-1,3-glucan-graft-poly(ε-caprolactone) (Glucan-g-PCL) (b)

Glucan-g-PCL copolymer was synthesized via classical ring-open polymerization (ROP) of N,N-dimethylamino-4-pyridine (DMAP) as an activator. Compared to other activators such as Sn(Oct)2, DMAP has shown higher catalytic activity because the distorted conformation of ε-caprolactone (ε-CL) promotes the favorable nucleophilic attack of the hydroxyl groups in α-1,3-glucan onto the carbonyl carbon, which is aided by the presence of DMAP [17, 26, 27]. To create a highly homogeneous ROP environment and enhance the accessibility of monomers to the inner core of the α-1,3-glucan backbone, the ionic liquid BmimCl, a much stronger hydrogen bond acceptor, was employed to completely dissolve α-1,3-glucan and ε-CL. Ionic liquids have been recognized as suitable green solvents for polysaccharide dissolution, aligning with the principles of green chemistry and contributing to the overall environmental sustainability of the synthesis process [15, 28, 29].

Characterization of Glucan-g-PCL copolymer

The FT-IR spectra of α-1,3-glucan and Glucan-g-PCL copolymer are shown in Fig. 2a. In the spectrum of α-1,3-glucan, the assignments for the characteristic peaks at 3275, 2922, and 922 cm−1 have been previously reported [15, 30]. The bands at 2934 cm−1 and 2868 cm−1 (C–H stretching), 1720 cm−1 (ester C=O group), and 1140 cm−1 (C–H bending) were present in the spectra of Glucan-g-PCL copolymer but absent in unmodified α-1,3-glucan, which confirmed the successful ROP of PCL with α-1,3-glucan [17, 31].

Fig. 2
figure 2

Characterization of α-1,3-glucan and Glucan-g-PCL copolymer. FT-IR (a), 1H NMR spectra (b), TGA curves (c), and XRD (d)

The 1H NMR spectra of α-1,3-glucan and Glucan-g-PCL copolymer are shown in Fig. 2b. The resonance peaks derived from the protons of the anhydroglucose unit (AGU) of the polysaccharide backbone appeared at δ = 3.19 ppm to 3.74 ppm; then, at δ = 4.60, 4.72, and 5.25 ppm, peaks corresponding to protons of residual hydroxyl groups in α-1,3-glucan can be observed [8, 32]. In the spectrum of Glucan-g-PCL copolymer, the signals of PCL methylene protons appeared at δ = 4.0 ppm (–CH2O–, e, repeating units), 3.37 ppm (–CH2OH–, e′, end unit), 2.26 ppm (–COCH2–, a, end unit), 1.53 ppm (–COCH2–, a, repeating unit), 1.40 ppm (–CH2–, b, d), and 1.30 ppm (–CH2–, c), respectively [17, 31, 33]. The 1H NMR spectral investigation presented in this experiment provides independent evidence for the successful graft reaction between α-1,3-glucan and PCL, which is consistent with the result of FT-IR spectra.

Based on these assignments, the microstructure of the Glucan-g-PCL copolymer, including the MS (molar substitution of PCL), DS (substitution degree of PCL chains, e.g., the number of PCL side chains per α-1,3-glucan unit), DP (polymerization degree of PCL), and WPCL (content of PCL) of the PCL side chains, was calculated using in the following equations [17, 31]:

$$\mathrm{MS}=\frac{{\text{CL}}}{{\text{AGU}}}=\frac{{I}_{({a+a}^{\mathrm{^{\prime}}})}/2}{{I}_{{({\text{O}}}_{2}{\text{H}}+{{\text{O}}}_{4}{\text{H}}+{{\text{O}}}_{6}{\text{H}})}/3}$$
$$\mathrm{DS}=\frac{\mathrm{terminal\; CL}}{{\text{AGU}}}=\frac{{I}_{a\mathrm{^{\prime}}}/2}{{I}_{{({\text{O}}}_{2}{\text{H}}+{{\text{O}}}_{4}{\text{H}}+{{\text{O}}}_{6}{\text{H}})}/3}$$
$$\mathrm{DP}=\frac{{\text{MS}}}{{\text{DS}}}=\frac{{I}_{a}}{{I}_{a\mathrm{^{\prime}}}}+1$$
$${\mathrm W}_{\mathrm{PCL}}=\frac{114{\text{MS}}}{162+114{\text{MS}}}\times 100\mathrm{\%}$$

where I is the peak integral (a: a′: O6H: O2H: O4H = 16.83:16.16:18.06:11.50:9.80), and 114 and 162 (g mol−1) in the expression for WPCL are the molecular weights of ε-CL monomer and AGU repeating unit, respectively. The calculated values for the structural factors of the Glucan-g-PCL copolymer were as follows: MS = 1.26, DS = 0.62, DP = 2.04, and WPCL = 46.99%.

The TGA and DTGA curves of α-1,3-glucan, Glucan-g-PCL copolymer are shown in Fig. 2c. The single thermal decomposition stage of α-1,3-glucan caused nearly 50% of weight loss at 319 °C, whereas the Glucan-g-PCL copolymer displayed the 50% weight loss temperatures at 306 °C. In the DTG curve, α-1,3-glucan showed one peak for the maximum decomposition rate at 321 °C, whereas the Glucan-g-PCL copolymer displayed a sharp peak for the maximum weight loss rate at 296 °C. Moreover, a minimal degradation peak in the 340–400 °C range was obtained and corresponds to the scission of PCL chains from the α-1,3-glucan backbone [31]. The lower thermal stability of Glucan-g-PCL copolymer was due to the lower crystallinity of α-1,3-glucan backbone and the introduction of PCL side chains [17, 33].

The XRD curves of the α-1,3-glucan and Glucan-g-PCL copolymer are shown in Fig. 2d.

The main diffraction signals of α-1,3-glucan appeared at approximately at 2θ = 10.64°, 14.62°, 19.59°, and 21.74°. However, chemical modification destroys the original crystallization of α-1,3-glucan by breaking the hydrogen bondings between α-1,3-glucan molecules; thus, all the mentioned peaks had disappeared and only one wide diffraction band at around 2θ = 20.15° was found in the XRD patterns of Glucan-g-PCL. The band at ∼20° is typically assigned to amorphous polysaccharide materials; these results are consistent with previous XRD reports that after the dissolution and chemical modification of α-1,3-glucan, cellulose, and other polysaccharides [15, 21, 29, 31].

Self-assembly behavior of Glucan-g-PCL copolymer

Upon the introduction of hydrophobic PCL side chains, the synthesized amphiphilic Glucan-g-PCL copolymer undergoes self-assembly in water, forming core-shell micelle-like nanostructure with PCL cores surrounded by α-1,3-glucan coronas. The particle size distribution and morphology of the resulting nanostructure were determined using DLS and TEM, and the results are displayed in Fig. 3a and b, respectively. The DLS results showed that the average particle size of Glucan-g-PCL nanostructure was 113.2 ± 4.29 nm and the size distribution (PDI) was 0.21 ± 0.05 with a monodispersed distribution. In this study, we used dialysis to prepare a self-assembled nanostructure. This approach can prevent the formation of micellar aggregates to a large extent as the system attains a thermodynamic equilibrium, thus mitigating micellar aggregation tendencies [5]. The morphology of the self-assembled nanostructure, as presented in the TEM image, consisted of individual nanoparticles with regular spherical shapes. Spherical micelle-like nanostructures have been reported to increase cellular uptake [34]. Furthermore, the particle size of nanostructure measured by TEM was smaller than that obtained by DLS; this may be attributed to the removal of water leading to shrinkage of hydrophilic α-1,3-glucan outer shell of nanostructure during the TEM sample preparation [35].

Fig. 3
figure 3

DLS analysis of size distribution (a), TEM image (b) of Glucan-g-PCL nanostructure, pyrene emission spectra of Glucan-g-PCL solutions with different concentrations (mg mL−1) (c), and intensity ratio (I1/I3) of the pyrene emission spectra versus of the logarithm concentration of Glucan-g-PCL copolymer (d)

When the concentration of the amphiphilic copolymer reaches a critical threshold known as the CAC, similar to traditional surfactants, a self-assembled nanostructure begins to form [36]. The CAC of the Glucan-g-PCL copolymer was determined using the fluorescent probe method. The photoluminescence of pyrene exhibits high sensitivity to alterations in the polarity of the surrounding environment. Given its hydrophobic characteristics, pyrene tends to position itself within the hydrophobic core of the nanostructure. Consequently, the photophysical properties of pyrene undergo notable changes, as illustrated in Fig. 3c. The intensity ratios of 373 nm and 384 nm I1/I3 in the emission spectrum of pyrene were plotted against the log concentrations of the Glucan-g-PCL copolymer in water (Fig. 3d). The CAC of the Glucan-g-PCL copolymer was calculated based on the crossover point, whereas I1/I3 began to decrease rapidly and was found to be 0.035 mg mL−1. The low CAC value observed in this study indicated that the Glucan-g-PCL nanostructure maintained their integrity even upon extreme dilution.

Encapsulation of Qu in Glucan-g-PCL nanostructure

We used self-assembled Glucan-g-PCL nanostructure as carriers to enhance the dispersibility of quercetin (Qu) in aqueous solutions. The Qu-loaded Glucan-g-PCL micelle-like nanostructure (Qu/M) was prepared using the sonication-emulsion solvent evaporation method. Upon dispersion in the Glucan-g-PCL nanostructure solution, Qu tended to precipitate in the hydrophobic domain within the nanostructure core and was thus physically encapsulated in the nanostructure (Fig. 4a). The results of UV-vis absorption spectra of blank nanostructure solution, Qu/water, and Qu/M also successfully validated the success of Qu’s encapsulation (Fig. S1). The Qu-loading parameters and particle sizes of Qu/M with different concentrations of Qu are shown in Table 1 and Fig. 4b and c. Figure 4b shows that, with an increase in Qu concentration, the loading content (LC, %) increased whereas the encapsulation efficiency (EE, %) decreased. Specifically, when the weight ratio of Qu to the copolymer increased from 1:20 to 1:1, the LC of Qu significantly increased from 3.33 to 9.65%. Conversely, the EE values exhibited a noticeable decline from 66.64 to 9.50%. It can be found that the Qu loading rate of the Glucan-g-PCL nanostructures can be controlled by adjusting the amount of Qu input and the nanostructures reach the equilibrium of Qu loading rate at 1:1 input. These results are consistent with those of previous studies and suggested that the Qu concentration is the main factor limiting the loading content during the initial phase of the loading process [21, 37].

Fig. 4
figure 4

Self-assembly of Glucan-g-PCL nanostructure in an aqueous solution (a). Encapsulation efficiency (EE) and loading content (LC) of Qu in Glucan-g-PCL nanostructure with different Qu feed ratio (b). Particle size and PDI of Qu-loaded Glucan-g-PCL nanostructure with different Qu feed ratio (c). DLS analysis of size distribution (d). TEM image of Qu/Glucan-g-PCL nanostructure (e) and chemical/physical characterization of Qu/Glucan-g-PCL nanostructure, XRD (f), and FT-IR (g)

Table 1 Characteristic parameters of blank nanostructure and Glucan-g-PCL nanostructure loaded with different Qu concentration

The particle size and solution properties of Qu/M with different Qu loading amounts were investigated using DLS. As presented in Fig. 4c, the particle size of Qu/M was significantly larger than that of the blank nanostructure, indicating successful encapsulation of Qu in the core of the Glucan-g-PCL nanostructure. Additionally, it was observed that the particle size of Qu/M increased from 147.8 to 201.4 nm as the weight ratio of Qu increased from 1:20 to 1:1, which were caused by the introduction of Qu inducing the changes in the inter- and intra-aggregate forces and their balance [38]. With the introduction of Qu, the PCL chains forming the core were more stretched and the interaction between α-1,3-glucan main chains forming the corona was reduced [21]. It has been reported that nanostructure smaller than 200 nm can efficiently deliver drugs to tumor cells through an enhanced permeability and retention (ERP) effect [39]. Considering both the size and Qu-loading capability, the Qu-loaded nanostructure with a weight ratio of 1:1 was chosen for subsequent studies.

Characterization of Qu-loaded Glucan-g-PCL nanostructure

The DLS result and TEM image of Qu/M are shown in Fig. 4d and e. Similar to the blank nanostructure solution, Qu/M also showed a spherical structure with a mono-dispersed size distribution. Furthermore, the crystallinity of the nanostructure, as well as the possible interaction between Qu and the Glucan-g-PCL copolymer in the nanostructure, was investigated using XRD and FT-IR. A crystallographic assay was performed using XRD, and the results are presented in Fig. 4f. The XRD pattern of the Qu powder showed specific peaks at 12.40° and 27.35°, which is in accordance with a previous study [24, 40]. In the XRD pattern of the physical mixture of Qu powder and the Glucan-g-PCL copolymer, the characteristic peaks of Qu were still observed with decreasing intensities, suggesting a crystalline state of Qu in the sample. In contrast, only a broad hollow peak pattern was observed for Qu/M, suggesting that Qu was no longer present in the crystalline state but in an amorphous state in the formulation. Moreover, the characteristic peaks of Qu disappeared and only the peaks of the Glucan-g-PCL copolymer were observed in the Qu-loaded Glucan-g-PCL nanostructure, further evidence that Qu was amorphous in the formulation. Similar observations of amorphous drugs (quercetin, curcumin, and diclofenac) in nanoparticles have been documented in several studies [19, 24, 41]. The FT-IR spectra (Fig. 4g) also clearly show the presence of Qu in the Glucan-g-PCL nanostructure. The spectra of Qu powder showed the characteristic peaks at a = 1657 cm−1, b = 1604 cm−1, and c = 1516 cm−1, attributed to the carbonyl (C=O) and aromatic rings, respectively [42]. As expected, when comparing the spectra of freeze-dried Qu/M with those of the physical mixture of Qu and Glucan-g-PCL copolymer powder, no new absorption peaks appeared, indicating the absence of any chemical reactions during the sample preparation procedures [19].

Antioxidant study and drug release in vitro

The antioxidant effect of the Qu encapsulated in the Glucan-g-PCL nanostructure was evaluated by measuring the number of DPPH radicals in the solution (Fig. 5a). The free radical scavenging activities from the untreated Qu powder, untreated α-1,3-glucan powder, Glucan-g-PCL nanostructure, Qu/M (1:20), Qu/M (1:10), Qu/M (1:5), and Qu/M (1:1) were 4.67% ± 0.53%, 3.82% ± 1.41%, 4.76% ± 2.25%, 33.37% ± 1.49%, 62.69% ± 3.04%, 78.11% ± 1.60%, and 93.17% ± 4.24% respectively. The untreated Qu powder showed a lower antiradical activity owing to its poor solubility. Glucan-g-PCL nanostructural encapsulation had a markedly enhanced antiradical activity compared with the untreated Qu powder, and the antiradical activity of Qu/M was up to 7–16 times greater than that of the untreated Qu powder. The hydrophobic core in the nanostructure formed by the PCL side chains provided a hydrophobic core for Qu encapsulation, which enhanced the water the solubility of Qu in aqueous solutions, allowing it to be preserved in the solution after filtration and increasing its availability in aqueous systems [43]. These results suggest that the encapsulation of Qu by using Glucan-g-PCL nanostructure is a useful way to enhance Qu dissolution and antiradical activity.

Fig. 5
figure 5

Antioxidant effect of (a) untreated Qu power, (b) α-1,3-glucan, (c) Glucan-g-PCL copolymer, (d) Qu/M (1:20), (e) Qu/M (1:10), (f) Qu/M (1:5), (g) Qu/M (1:1) (a). Cumulative release of Qu from Glucan-g-PCL nanostructure at pH 7.4 PBS (0.02 M) with 1% Tween-80 (b). Storage stability of Glucan-g-PCL nanostructure and Qu-loaded Glucan-g-PCL nanostructure at 4 °C (c). Degradation curve of Qu-loaded Glucan-g-PCL nanostructure during storage at 4 °C (d)

To evaluate Qu release profiles from the Qu-loaded Glucan-g-PCL nanostructure, in vitro release study was carried out in a pH 7.4 PBS at 37 °C using the Qu PBS solution as the control. As depicted in Fig. 5b, the Qu PBS solution showed a large initial burst release, with approximately 89% total release in the first hour. For the Qu-loaded Glucan-g-PCL nanostructure, a two-stage release pattern was observed: an initial burst release of approximately 32.39% of Qu in the first hour, followed by a moderately controlled release in the subsequent 7 h and with a cumulative release rate of 66.67%. The initial burst release may be due to ultrasonic treatment resulting in the deposition of Qu onto the surface of Glucan-g-PCL nanostructure or at the interface between the nanostructured core and the corona being transferred to the medium [15, 19, 37]. Such a burst release, especially for polysaccharide-based nanostructure, could be observed in many other reports that use the after-loading method (ultrasonic treatment of nanoclusters/nanostructures) [15, 22, 23, 37]. Different loading methods (physical or chemical), which may alter the drug release trend from the nanostructures, which are suggested for further investigations. Notably, approximately 33% of Qu was still left in the Qu/M after 8 h, which may be due to the increased Qu concentration in the PBS medium affecting its diffusion [37]. In addition, to study the kinetics of drug release, four types of commonly used kinetic models—zero-order, first-order, Higuchi, and Ritger-Peppas—were applied, and the kinetic model was determined by comparing the curve-fitting method and their correlation coefficient values [44]. Table 2 shows that the first-level model was the most consistent with the kinetic data of Qu release, indicating that the main mechanism of Qu release from Glucan-g-PCL nanostructure was diffusion, that is, random molecular motion driven by the Qu concentration gradient [41, 45].

Table 2 Kinetics parameters of Qu release from the prepared nanostructure

Stability of Qu-loaded Glucan-g-PCL nanostructure

The investigation into its physical stability in this study (Fig. 5c) demonstrated a slight increase in the average particle size and PDI of blank nanostructure during a 4-day storage period at 4 °C. However, the system subsequently achieved a state of relative stability, and the observed increase in nanostructure diameter could potentially be attributed to nanostructure aggregation [5, 15]. In the case of Qu/M, during the first 8 days, unlike the blank nanostructure, a slight decrease in the average particle size was observed. This trend could be associated with the release or degradation of encapsulated Qu. To validate this hypothesis, the degradation rate of Qu confined within the nanostructure during storage was assessed. The retention rate of Qu loaded into nanostructure during storage is depicted in Fig. 5d. Overall, a declining trend was observed as the storage time increased, with 18.01% of quercetin lost after being stored at 4 °C for 25 days.

Conclusion

Water-insoluble α-1,3-glucan was prepared through enzymatic polymerization with GTF-I enzyme from S. mutans ATCC 700610, and a novel enzymatic polymerization-derived α-1,3-glucan-based amphiphilic copolymer, Glucan-g-PCL, was successfully synthesized by grafting modification of α-1,3-glucan and ε-caprolactone. The synthesized amphiphilic α-1,3-glucan copolymer forms micelle-like nanostructure by self-assembly in water and showed high dilution stability with a very low CAC value (35 mg L−1). Quercetin, a poorly water-soluble model compound, was successfully encapsulated in the core of Glucan-g-PCL nanostructure using a simple sonication-emulsion solvent evaporation method, and it was present in an amorphous state in the formulation. Furthermore, the Qu-loaded nanostructure formed by the copolymer in PBS showed sustained drug release; it would be desirable for achieving therapeutic drug level for a prolonged period with relatively less side effects [46]. Moreover, the antioxidant activity of the Qu-loaded nanostructure, investigated by the DPPH method, was effectively enhanced in comparison with Qu/water. Both blank nanostructure and Qu/M solution showed good storage stability within 1 month. In conclusion, amphiphilic α-1,3-glucan copolymer, Glucan-g-PCL, and its self-assembled nanostructure can serve as promising nanocarriers for the efficient loading and delivery of hydrophobic drugs, but further in vitro and in vivo studies are still needed to explore their efficacy, cytotoxicity, biocompatibility, and degradability.