β-cyclodextrin/alginate nanoparticles encapsulated 5-fluorouracil as an effective and safe anticancer drug delivery system

2.1 Materials

Materials (calcium chloride (CaCl₂), β-cyclodextrin and sodium alginate) were purchased from Shanghai Chemical Co. (China). 5-FU and sulforhodamine B (SRB) were provided by Sigma Co. (USA). Trichloroacetic acid and acetic acid were purchased from Merck (Germany). All chemicals were used without further purification. Deionized water was used in all experiments.

2.2 Fabrication of 5-FU encapsulated β-CD/Alg nanoparticles

The synthesis of β-CD/Alg nanocomposites was performed via the ionotropic gelation mechanism previously reported, with slight modification (Nguyen et al., 2018, Nguyen et al., 2019a). Briefly, different concentrations of 5-FU solution were added dropwise to 5 mL of CaCl₂ solution (6.5 mg/mL) and stirred at 1200 rpm for 90 min. Then, the mixture was added to 13.7 mL of Alg solution (10 mg/mL), stirred for 90 min, and irradiated with ultrasound for 60 min. The 5-FU@Ca/Alg mixture was centrifugated and washed with water (3 × 10 mL) before the addition of the β-CD solution (10.7 mL, 1.6 mg/mL). The mixture was stirred then left for 24 h, centrifugated at 3200 rpm for 15 min and washed with water (3 × 10 mL). The nanocomposites were obtained via a freeze-drying process after 14 h.

2.3 Association efficiency and loading efficiency

The association efficiency of the drug is defined as the ratio of 5-FU by weight in the nanocomposite to the weight of the nanocomposite obtained, as shown in Eq. (1). The weight of 5-FU in the nanocomposite was estimated from the difference between the total amount of 5-FU (W_{total 5-FU}) and the amount of 5-FU in the supernatant separated from the nanocomposite after centrifugation (W_free 5-FU). The amount of the drug in the supernatant was determined via a calibration curve of absorption intensity at a peak of 266 nm in UV–Vis measurement (Yusefi et al., 2021).

The loading efficiency of the drug is the ratio of 5-FU by weight in the nanocomposite to the weight of the dry nanocomposite, obtained via Eq. (2).

2.4 Physicochemical characterisation of the nanocomposite

Zeta potential was used to determine the electrochemical equilibrium at the interfaces and molecular vibrations, while DLS was used to determine the particle size distribution in the colloidal solution. The nanocomposite was diluted in deionized water to attain a solution with a concentration of 0.5 mg/mL. A nanoPartica Horiba SZ-100 (Japan) was used for the measurements. Zeta potential was measured at 25 °C with an applied voltage of 3.3 V, and DLS was performed at an angle of 173°.

To determine the presence of possible functional groups in the synthesized nanocomposite, the samples including blank β-CD/Alg and 5-FU@β-CD/Alg were characterized by Fourier-transform infrared (FTIR) analysis. The measurement was carried out on a Tensor 27 FTIR spectrophotometer (Brucker, Germany) with a wavenumber range of 500–4000 cm⁻¹ and resolution of 0.5 cm⁻¹.

The X-ray diffraction (XRD) patterns of 5-FU, β-CD/Alg and 5-FU@β-CD/Alg crystals was measured on an X-ray diffractometer (Bruker, Model-D8 Advance).

The particle size distribution and morphology of the nanocomposite 5-FU@β-CD/Alg were evaluated using TEM (S-4800 JEOL JEM1400) set at an accelerating voltage of 120 kV and FESEM (S-4800 HI-9057–0006).

Thermogravimetry analysis (TGA) and differential scanning calorimetry (DSC) were performed using a LabSys Evo 1600 thermal analyser (SETARAM, France) at a temperature range of 30–800 °C and a heating rate of 10 °C min⁻¹ in air.

2.5 In vitro release profile and kinetic evaluation of 5-fluorouracil

A drug release study was performed in simulated colorectal fluid (phosphate-buffered saline at pH 7.4) and gastric fluid (hydrochloride/potassium chloride buffer at pH 1.2). The drug release efficiency of 5-FU@β-CD/Alg was evaluated using a 5 mL-dialysis bag (cut-off molecular weight between 3000 and 5000 Da) as previously reported, with slight modification (Fu and Kao 2010, Yusefi et al., 2020). Before the experiment, the dialysis bags were soaked in a release medium for 12 h. The solution mixture comprising the nanocomposite 5-FU@β-CD/Alg (3 mg) and the release medium (3 mL) was added to the dialysis bag with the two ends tied. The dialysis bag was immersed in 12 mL of the release medium maintained at 37 °C and slowly stirred. To detect the released drug, an aliquot (0.5 mL) was withdrawn at different times in range of 0.25–96 h and the outer solution was immediately replaced by the same volume of fresh medium. The withdrawn samples were characterized by UV–Vis spectroscopy with a wavelength of 266 nm. Runs were performed in triplicate. The release profile was plotted as the relative release percentages of 5-FU over time. The same procedure was performed for blank β-CD/Alg and 5-FU as standard references. The drug release from the different media was calculated via Eq. (3).

For release kinetic study, the following models including zero order (Eq. (4)), first order (Eq. (5)), Hixson-Crowell (Eq. (6)), Higuchi (Eq. (7)) and Korsmeyer-Peppas (Eq. (8)) were investigated via an origin software. The mathematical equations representing these models is listed as follows:

2.6 Cytotoxicity assay

Fibroblast and MCF-7 (HTB-22) cells were provided and cultured by the University of Science, Vietnam National University, Ho Chi Minh City, Vietnam. Cells were cultured at 37 °C and 5% CO₂ in Eagle’s minimal essential medium (EMEM) supplemented with 20 mM HEPES (Sigma), 0.025 μg/mL amphotericin B (Sigma), 2 mM L-glutamine (Sigma), 100 IU/mL penicillin G (Sigma), 100 μg/mL streptomycin (Sigma), and 10 % (v/v) FBS (Sigma). Samples with β-CD/Alg, 5-FU@β-CD/Alg and the solutions after release at pH 7.4 and pH 1.2 were used to test cytotoxicity.

The SRB assay was carried out as previously reported, with some modification (Nguyen and Ho-Huynh 2016, Thach et al., 2021). Cells were seeded in 96-well plates with 10,000 cells/well and cultured for 24 h before being incubated with the samples at various concentrations for 48 h. The cells were placed in cold trichloroacetic acid solution (50 %, w/v) for 1–3 h, washed, and stained with SRB (0.2%, w/v) for 20 min. Water and camptothecin (Calbiochem) were used as negative and positive controls, respectively. Absorption intensity was determined using an ELISA plate reader (Synergy HT, Biotek Instruments) at wavelengths of 492 nm and 620 nm. The percentage of growth inhibition (%) was determined at a concentration of 100 μg/mL via Eq. (9).

3.1 Synthesis of nanocomposite 5-FU@β-CD/Alg

Synthesis of nanocomposite 5-FU@β-CD/Alg was carried out via an ionotropic gelation mechanism using Ca²⁺/Alg gelispheres and β-CD molecules as previously reported, with slight modification (Nguyen et al., 2015, Nguyen et al., 2019a). In the present work, the gelispheres were loaded with 5-FU before the nanoparticles were synthesised, as illustrated in Fig. 1. β-CD molecules have a negative zeta potential and can interact to cross-link a Ca²⁺ ion and alginate in the insoluble gelispheres to form a nanocomposite that is well dispersed in water. 5-FU can be encapsulated in the matrix of the polysaccharides via bonds with Ca²⁺ ions and functional groups in the CD molecules and alginate chain. The nanocomposite 5-FU@β-CD/Alg is easily purified and obtained using centrifugation. Because the quantity of the drug can significantly influence the efficiency of nanocomposite synthesis, the loading process for the drug was investigated using different ratios of drug weight to carrier weight. The loading efficiency and association efficiency of the drug were used to evaluate the performance of synthesis. The best nanocomposite was used to investigate the physicochemical characterisation and in vitro drug release behaviour and kinetics.

Close

Fig. 1

Study scheme and proposed formation mechanism for the 5-FU@β-CD/Alg nanocomposite.

Export to PPT

To fabricate the nanocomposite, the same procedure as a previously reported was used, with a hydrophobic drug (ketoprofen) loaded onto β-CD before the 5-FU@β-CD complex was added to the Ca²⁺/Alg galispheres to form the nanocomposite (Nguyen et al., 2015). However, only a very low loading efficiency of the 5-FU was obtained and no nanoparticles were formed in the solution; this is likely because the solubility of 5-FU is greater than ketoprofen and its molecular size is smaller. Recent research (Jin et al., 2010, Di Donato et al., 2016) has confirmed that β-CD can easily encapsulate 5-FU in the solid state. However, it is not a good host for the complexation of 5-FU in an aqueous solution due to the low stability of the complex (Melnikova et al., 2020). Thus, 5-FU can interrupt the interaction between Ca²⁺/Alg galispheres and β-CD molecules during nanoparticle formation in solution, impeding nanocomposite synthesis.

In recent research (Nguyen, Vo, Huynh, et al. 2019), nanoparticles have been fabricated with changes in the synthesis procedure and a different order of addition of materials. With this change, 5-FU was entrapped into Ca²⁺/Alg galispheres before β-CD molecules were added to form the nanoparticles (Fig. 1); this procedure performed well in synthesising the 5-FU@β-CD/Alg nanocomposite. To study the effect of drug concentration, feeding ratios of 5-FU in the range 5–40% (w/w) were used to determine drug association efficiency and drug loading efficiency. The results show a non-linear relationship between the drug content added to the nanocomposite and encapsulation and loading efficiency (Fig. 2). Raising feeding ratios of 5-FU from 5 % to 40 % decreases the association efficiency of the drug from 96.8 ± 1.5% to 81.7 ± 2.0%, due to the saturation of 5-FU in the matrix of alginate chains. Meanwhile, increase in these ratios increases the loading efficiency, with a maximum of 36.4 ± 1.7 % at a feeding ratio of 20% (w/w). Although a non-linear relationship between feeding ratios and encapsulated efficiency can be observed for the β-CD-based nanocomposite, this result represents a different trend for 5-FU encapsulation than for other nanocomposites reported previously. Indeed, Prabha and Raj (Prabha and Raj 2016) showed that the association and loading efficiencies of 5-FU from the nanocomposite β-CD/Fe₃O₄/PEG increase with an increase in 5-FU concentration. Similarly, Hosseinifar et al. (Hosseinifar et al., 2018) also found that an increase in drug ratios induces better encapsulation performance in β-CD-Alginate nanogel.

Close

Fig. 2

Influence of 5-FU concentration on loading efficiency and association efficiency of the drug from the 5-FU@β-CD/Alg nanocomposite.

Export to PPT

3.2 Physicochemical characterisation of the nanocomposite 5-FU@β-CD/Alg

Zeta potential and particle size are the principal criteria for evaluating effective anticancer drug delivery targeting tumour tissue. Fig. 3 shows spectra of the zeta potential and particle size of blank nanocomposite β-CD/Alg and nanocomposite 5-FU@β-CD/Alg. Because of the negative zeta potential of the components β-CD and Alg (Lakkakula et al., 2017), the blank nanocomposite β-CD/Alg has a great surface charge of −48.7 mV while the presence of 5-FU (20%) slightly decreases the surface charge (−42.1 mV). The highly negative zeta potential induces repulsion between nanoparticles, greatly reducing the propensity to aggregate and making the nanocomposite highly stable in an aqueous solution. Thus, the nanocomposite based on β-CD and Alg protects the drug and is an ideal carrier for a drug delivery system.

Close

Fig. 3

Zeta potential, dynamic light scattering and FESEM images of an aqueous solution of β-CD/Alg (A, C & E) and 5-FU@β-CD/Alg nanocomposite (B, D & F).

Export to PPT

The particle size of the nanocomposites in an aqueous solution was determined by DLS measurement. The result shows that the particle size of 5-FU@β-CD/Alg is in the range 20–200 nm with a mean size of 57 nm, while the particle size of β-CD/Alg is larger, at 100–250 nm with a mean size of 150 nm. The smaller size of the nanocomposite 5-FU@β-CD/Alg indicates that the presence of 5-FU may significantly reduce the size of the nanoparticle, confirming the encapsulation of 5-FU into the matrix of crosslinks formed by calcium ions and alginate chains in β-CD/Alg. The FESEM images (Fig. 3E and 3F) show that the nanocomposites possess regular fibres with sheet-like shapes which were formed during the drying process.

The FTIR spectra of blank β-CD/Alg, 5-FU and 5-FU@β-CD/Alg nanocomposite are shown in Fig. 4. The FTIR spectrum of blank β-CD/Alg has the characteristic absorption peaks of glucose units at 3425, 2925, 1418 and 1030 cm⁻¹. The bands at 3425 cm⁻¹ and 1418 cm⁻¹ can be attributed to the stretching and bending vibrations, respectively, of OH groups and the band at 1030 cm⁻¹ to the stretching vibration of glucose C-O-C groups. The band at 2925 cm⁻¹ is characteristic of C-H stretching vibration. In particular, the peak at 1627 cm⁻¹ corresponds to the stretching vibration of the carboxyl groups (COO) in alginate chains. The FTIR spectrum of 5-FU has bands at 3142, 2936, 1776, 1349 and 1249 cm⁻¹, where the bands at 3142 cm⁻¹ and 2936 cm⁻¹ are characteristic of N-H and C—H stretching vibrations, respectively, the band at 1776 cm⁻¹ is characteristic of —C⚌O group of ketone, and the bands at 1349 and 1249 cm⁻¹ are characteristic of stretching vibration of C—N and C—F, respectively. Analysis of the FTIR spectrum of the 5-FU@β-CD/Alg nanocomposite shows that the absorption bands shift to new positions compared to the blank nanocomposite, namely 3413, 2934, 1613, 1424, 1034 cm⁻¹. The broad band at 3413 cm⁻¹ can be assigned to the stretching vibration of N—H and O—H groups, the band at 2934 cm⁻¹ to the stretching vibration of C—H, and the bands at 1613 cm⁻¹ and 1424 cm⁻¹ to the stretching vibration of C⚌O groups and bending vibration of the OH groups, respectively. The band at 1034 cm⁻¹ is characteristic of the stretching vibration of C—O—C groups. Notably, the spectrum has bands from 1350 to 1250 cm⁻¹ corresponding to the characteristic vibrations of C—N and C—F groups in the 5-FU molecule. The results indicate that the 5-FU drug is successfully encapsulated into the carrier β-CD/Alg.

Close

Fig. 4

FTIR Spectra of (a) β-CD/Alg, (b) 5-FU and (c) 5-FU@β-CD/Alg nanocomposite.

Export to PPT

XRD patterns of 5-FU, β-CD/Alg and 5-FU@β-CD/Alg nanocomposite are shown in Fig. 5. The XRD pattern of 5-FU showed the intense peaks at 2θ value of 16.5 and 28.6° that relates to its crystalline nature. The data of β-CD/Alg nanocomposite shows an abroad diffraction peak appeared at 2θ value of 43.5⁰ which relates to crystalline structure glucuronate and mannuronate units in alginate chains with a crystalline diameter of 1.47 nm. For the diffractogram XRD of 5-FU@β-CD/Alg nanocomposite, the peaks observed at 16.2, 28.5 and 43.4⁰ indicates the successful loading of 5-FU into the matrix of β-CD/Alg composite and maintaining crystalline nature of drug and the polysaccharides. The crystal diameter calculated at peak 43.4⁰ was determined to be 1.36 nm.

Close

Fig. 5

XRD patterns of (a) 5-FU, (b) β-CD/Alg and (c) 5-FU@β-CD/Alg nanocomposite.

Export to PPT

The size and morphology of the 5-FU@β-CD/Alg nanocomposite were determined using TEM. Images and particle size distribution are shown in Fig. 6. The TEM images indicate that the nanocomposite has a uniform spherical morphology with low aggregation. The nanoparticles distribute in the range 30–80 nm with a mean diameter of 50 nm. The result confirms the successful synthesis of the 5-FU@β-CD/Alg nanocomposite.

Close

Fig. 6

Transmission electron micrograph images at different amplification and particle size distribution of 5-FU@β-CD/Alg nanocomposite.

Export to PPT

The thermal stability of the nanocomposites was analysed using TGA and DSC (Fig. 7). The TGA curves of both nanocomposites show three main weight-loss steps. The first weight loss step in both samples occurs between 30 and 220 °C and is estimated to be 22%. This step reflects the evaporation of adsorbed moisture from the surface of the samples, as confirmed by the presence of a corresponding endothermic peak in the DSC curves. The second weight-loss steps for β-CD/Alg and 5-FU@β-CD/Alg occur in the temperature range 220–545 °C with a weight loss of 43% and 220–440 °C with a weight loss of 40%, respectively. This weight loss can be attributed to the oxidation of sugar molecules in polysaccharide chains which relate to the small exothermic peaks in the DSC curves. In the final step, β-CD/Alg shows a weight loss of 20% in the range 545–800 °C and 5-FU@β-CD/Alg shows a weight loss of 25% in the much lower range 440–800 °C. The DSC curves have intense exothermic peaks in the corresponding temperature regions that can be attributed to oxidation of the drug and polysaccharide components. Moreover, enthalpy value of the drug loaded nanocomposite was found to be greater than the value of the blank nanocomposite that indicated the interaction between drug molecules and polysaccharides in the nanocomposite. The low-temperature stability of the 5-FU@β-CD/Alg nanocomposite indicates the presence of the drug in the nanocomposite. On the other hand, 5-FU is oxidised at a much low temperature, in the range 250–350 °C, as previously reported (Obireddy and Lai 2021). Thus, the thermal behaviour of the 5-FU@β-CD/Alg nanocomposite indicates the interaction of the drug with sugar molecules in polysaccharides that leads to a significant increase in the thermal stability of 5-FU.

Close

Fig. 7

TGA and DSC curves of β-CD/Alg and 5-FU@β-CD/Alg nanocomposite in airflow of 20 mL/min at a heating rate of 10 °C/min.

Export to PPT

3.3 In vitro drug release of nanocomposite 5-FU@β-CD/Alg

The drug release profile of 5-FU from the 5-FU@β-CD/Alg nanocomposite was investigated in different media at a constant temperature (37 °C). The 5-FU drug without a carrier in a phosphate buffer (pH 7.4) was also explored as a reference, with 5-FU release of 92.0% and 98.7% at 2 h and 4 h, respectively. Fig. 8A shows release profiles of the nanocomposite in phosphate buffer solutions at different pH values, including simulated colorectal fluid medium (pH 7.4) and simulated gastric fluid medium (pH 1.2) for 24 h. It was found that rapid release occurs during the first 60 min, followed by slower release. In the colorectal fluid medium (pH 7.4), drug release of 37.7% and 52.1% was obtained after 1 h and 4 h, respectively, with 69.8% after 24 h and 80.7% after 96 h. A much slower rate of release was observed in pH 1.2, with 20.0% released in the first 4 h and 34.1% in 96 h. Thus, the results indicate that the 5-FU drug in the nanocomposite is well protected from the physiological environment, and release behaviour from the nanocomposite is modulated by changes in environmental pH media, meaning that the nanocomposite can be used effectively in the targeted treatment of cancer cells.

Close

Fig. 8

Drug release profile of the 5-FU@β-CD/Alg nanocomposite at pH 7.4 and pH 1.2 (A); TEM images of the 5-FU@β-CD/Alg nanocomposite recovered after 48 h from pH 1.2 (B) and pH 7.4 (C).

Export to PPT

It is notable that the drug release behaviour from the 5-FU@β-CD/Alg nanocomposite is similar to previous reports on alginate-based nanocomposites (Azhar and Olad 2014, Anirudhan and Nair 2019, Wu et al., 2019), which have shown that at pH 7.4, 5-FU is released more rapidly than it is at pH 1.2. However, the release of the 5-FU drug is different from the release behaviour of ketoprofen from the nanocomposite ketoprofen@β-CD/Alg reported previously, where a lower pH leads to a faster drug release. This phenomenon demonstrates that 5-FU is loaded onto the matrix of Ca²⁺/Alg but is not forming an inclusion complex with β-CD.

In order to understand 5FU@β-CD/Alg release behaviour, the samples were recovered after release in both media and further characterised by TEM. The TEM images are shown in Fig. 8B and C. It is clear that the nanoparticles are clustered at pH 1.2 while the composite is sufficiently swelled at pH 7.4 to facilitate the diffusion of the drug into the solution. The formation of clusters at pH 1.2 contributes to the protonation of COO^– groups in alginate, leading to the recombination of particles to form clusters of greater size and network holes to be minified, restricting pervasion of 5-FU from the interior of the nanocomposite. It can be concluded that the release behaviour of the drug from the 5-FU@CD/Alg nanocomposite is mainly by way of changes in the morphology of the nanocomposite.

The release kinetics of 5-FU drug in the various media were modelled into the five principal mathematical models. The kinetic parameters are listed in Table 1. The drug release behaviour fits both Hugichi and Korsmeyer-Peppas models but the later release model is best fitted. The results show that the release of 5-FU from the nanocomposite at pH 1.2 possesses slightly higher coefficient in comparison with release behaviour at pH 7.4. However, the difference is not significant, thus it can be concluded that the release kinetic is well fitted in both media. In addition, for all environmental pH values, the diffusion exponent is found to be lower than 0.45, indicating that the 5-FU release from the nanocomposite follows Fickian diffusion mechanism.

3.4 Cytotoxicity assay

The cytotoxic effects of β-CD/Alg, 5-FU@β-CD/Alg and release solutions at different pH values after 48 h against a normal cell line (fibroblast) and a breast cancer cell line (MCF-7) are shown in Table 2. As can be seen, the β-CD/Alg nanocomposite displays no inhibition of the normal cell line but a slight inhibition of 26.83 ± 3.58 against the cancer cell line at a concentration of 100 μg/mL. However, IC₅₀ values of the blank nanocomposite are over 100 μg/mL for both cell lines, demonstrating the high biocompatibility of the carrier material. The results also show that although the IC₅₀ value of 5-FU against fibroblasts is >100 µg/mL, inhibition against this normal cell of about 45.31 ± 5.52% at a concentration of 100 µg/mL is much higher than inhibition with 5-FU@β-CD/Alg. This result reveals the drawbacks of using the drug alone in chemotherapy treatment without a drug delivery system. As expected, the 5-FU@β-CD/Alg nanocomposite shows high toxicity against the cancer cell lines but almost no toxicity against normal cells. Indeed, this nanocomposite inhibits 71.75 ± 3.73% and 20.59 ± 3.07% for MCF-7 and fibroblasts cells, respectively, while the IC₅₀ values are 21.68 ± 2.22 μg/mL and over 100 μg/mL, respectively. These results indicate that the nanocomposite is safe for normal cells but has a good ability against cancer cells. To evaluate the efficiency of drug release from the nanocomposite, the toxicity of the solution after release at pH 7.4 and pH 1.2 after 48 h was tested for both cell lines. Good toxicity of these release solutions was found against MCF-7, with inhibition of 78.48 ± 0.97% and 70.32 ± 0.98 %, respectively, and IC₅₀ values of 2.02 ± 0.38 and 10.96 ± 0.69, respectively, while this solution shows very low toxicity against normal cells. Additionally, statistical analysis of the data showed that IC₅₀ value of the 5-FU@β-CD/Alg nanocomposite released at pH 7.4 against MCF-7 is similar to the alone 5-FU drug but significantly different from the solution released at pH 1.2. Therefore, the results suggest the safety and high potential of using β-CD/Alg as an anticancer drug delivery system with high antiproliferation activity against MCF-7 cells and negligible effects on normal healthy cells.