3.1. Physicochemical Characterization of PCL NPs
The average particle size, polydispersity index (PDI), and zeta potential measurements were conducted for PCL placebo NPs produced by different manufacturing techniques. Dynamic Light Scattering (DLS) and Zeta Potential measurements were performed for each sample, and therefore PCL NPs-0 are summarized in Table 2.
The smallest diameter of NPs was achieved with 14000 g/mol PCL using the ultrasonic homogenizer after filtration through a 0.22 µm filter, yielding a particle size of 202.4 nm. The nanoparticle sizes exhibited no substantial change after filtration; therefore, the filtration step was eliminated from the manufacturing process. The ideal nanoparticle size for oral administration falls within the range of 100 nm to 800 nm, with particles smaller than 100 nm being suitable for parenteral administration. Conversely, larger particles exceeding 1000 nm may pose a risk of phagocytosis [15]. Consequently, it was concluded that nanoparticles obtained through both Ultra-Turrax and ultrasonic homogenizer methods were suitable for oral use.
The PDI represents the particle size distribution, and the PDI values of particles generated by ultrasonication were found to be less than 0.20. For the polymer-based NPs, PDI values below 0.2 are considered to be favorable [16]. According to the PDI results, NPs produced with ultrasonic homogenization demonstrate good uniformity and homogeneity in terms of particle size and distribution. However, in the case of processes conducted using Ultra-Turrax, PDI values were higher than 0.20, thus not deemed suitable in terms of uniformity and homogeneity.
Javaid et al. reported a zeta potential of -16.8 mV for placebo PCL NPs and − 11.1 mV for sefotaxime-loaded PCL NPs [17]. Mahmoudi et al. declared that placebo PCL NPs had zeta potential − 29.3 mV and memantine-loaded PCL NPs had − 26.5 mV zeta potential [18]. Diyanat et al. found that placebo PCL NPs had − 25.8 mV and pretilachlor-loaded PCL NPs had − 23.0 mV zeta potentials [19]. Based on these results, it can be inferred that as the hydrophobicity of the active substance to be encapsulated in PCL increases, the absolute zeta potential value decreases.
Based on the evaluation of the results, considering both size and stability, as well as standardization aspects, the most suitable production condition was achieved using 14000 g/mol PCL by ultrasonication at a 72% amplitude. Collectively, CBD encapsulation studies were proceeded with these conditions.
Table 2
The size, PDI and zeta potential for PCL NPs-0 (n = 3)
PCL (g/mol) | Homogenizer | rpm/ %Amp | Mean diameter | PDI | Zeta Potential (mV) |
14000 | Ultra-turrax | 25000 | 327.9 ± 6.51 | 0.259 ± 0.027 | -14.98 |
45000 | Ultra-turrax | 25000 | 338.1 ± 4.46 | 0.271 ± 0.038 | -17.17 |
80000 | Ultra-turrax | 25000 | 392.4 ± 13.79 | 0.250 ± 0.011 | -15.72 |
14000 | Ultrasonic | 72% | 207.5 ± 0.45 | 0.114 ± 0.044 | -23.37 |
14000a | Ultrasonic | 72% | 202.4 ± 6.36 | 0.110 ± 0.018 | -20.16 |
45000 | Ultrasonic | 72% | 235.6 ± 2.29 | 0.107 ± 0.020 | -21.12 |
80000 | Ultrasonic | 72% | 264.6 ± 6.85 | 0.108 ± 0.020 | -24.11 |
a) Filtered through a 0.22 µm filter.
3.1.1. Particle size, polydispersity index, and zeta potential of NPs
Average particle size, PDI and zeta potential for CBD-PCL NPs loaded with different amounts of CBD are presented in Table 3. For the CBD-loaded PCL NPs (5 mg, 10 mg, 15 mg, and 30 mg CBD), the zeta potential values were determined as -10.21, -9.62, -9.93, and − 9.40 mV, respectively. These values may indicate the presence of CBD partially on the particle surface despite being trapped within the polymer matrix, suggesting partial adsorption of CBD on the PCL shell and its incorporation with PCL. Consequently, it can be suggested that CBD increases the zeta potential in a positive direction by enhancing the surface charge of PCL NPs. Zeta potential values between 0 and ± 5 mV indicate a tendency of particles to aggregate or come together, leading to rapid particle aggregation. Particles between ± 5 and ± 20 mV are minimally stable, while particles between ± 20 and 40 mV are moderately stable [20]. Zeta potential values higher than ± 30 mV can increase stability but may also lead to toxicity [15]. NPs produced by the ultrasonic homogenizer exhibit moderate stability, while CBD-PCL NPs demonstrate minimal stability. Collectively, both placebo and CBD-PCL NPs exhibited zeta potential values below the risk threshold for toxicity.
Table 3
Mean diameter, PDI, and zeta potential for CBD-loaded PCL NPs, (n = 3)a
NPs | Mean diameter (nm) | PDI | Zeta Potential (mV) |
PCL NPs-1 | 227.2 ± 2.35 | 0.100 ± 0.004 | -10.21 |
PCL NPs-2 | 225.5 ± 0.53 | 0.133 ± 0.029 | -9.62 |
PCL NPs-3 | 227.7 ± 1.55 | 0.094 ± 0.014 | -9.93 |
PCL NPs-4 | 216.9 ± 6.81 | 0.126 ± 0.031 | -9.40 |
a) Ultrasonic homogenizer was employed with 72% amplitude.
3.1.2. FTIR analysis
FT-IR analysis was conducted to investigate the molecular interaction between the active ingredient and the polymer. The characteristic bands of pure PCL and pure CBD are presented in Fig. 1, while the FT-IR spectra of pure PCL, pure CBD, placebo PCL NPs (PCL NPs-0), and CBD-loaded PCL NPs are revealed in Fig. 2.
Characteristic bands of CBD were identified at 3518 cm-1, 3404 cm-1, 1628 cm-1, and 1585 cm-1 [21]. In the FT-IR spectrum of CBD, the prominent bands in the range of 3404 cm-1 to 3518 cm-1 corresponded to O-H stretching vibrations, while the bands at 3074 cm-1 represented C-H stretching (phenyl), the band at 2924 cm-1 indicated methyl and methylene groups, the band at 1585 cm-1 represented C = C stretching in the phenyl ring, and C-O stretching vibrations were observed at 1213 cm-1 [22].
Characteristic bands of PCL were observed at 2866 cm-1 for symmetric aliphatic stretching and at 2943 cm-1 for asymmetric aliphatic stretching. The strong characteristic carbonyl (C = O) group was observed at the stretching band of 1720 cm-1. The C-O and C-C stretching vibration bands of PCL were observed at 1292 cm-1. Additionally, the symmetric C-O-C vibration band was observed at 1168 cm-1 and the asymmetric C-O-C vibration band at 1238 cm-1 [8].
The FT-IR spectrum of PCL NPs-4 exhibited characteristic bands at 1585 cm-1 and 1628 cm-1 corresponding to the chemical structure of CBD. The stretching vibration caused by the OH group was present in all CBD-loaded PCL NPs at bands 2860 cm-1 and 2945 cm-1. The band at 1442 cm-1, attributed to the deformation of methylene or methoxy groups (C-H deformation), cannot be considered as a representative band for CBD [23]. The presence of CBD within PCL can be emphasized by comparing the PCL bands with the presence of bands belonging to aliphatic groups (= C-H) at 1585 cm-1 and 1628 cm-1. These two specific bands have also been observed in previous study of Andriotis et al. in their CBD/cyclodextrin formulation and cannabidiol orodispersible tablets formulated by Vlad et al. [23, 24]. In this work, the peaks of CBD were observed in PCL NPs-4 sample more profoundly than in PCL NPs-1, 2, 3. This may be due to the molecular dispersion of low CBD dose within the PCL polymer matrix which was also previously observed by Nawaz et al in their studies with PCL encapsulated acriflavine [10]. To conclude, FTIR spectra demonstrates CBD and PCL exhibited no chemical incompatibility in NPs.
3.2. Morphology of NPs
The morphological characteristics of PCL NPs-0 and PCL NPs-4 were examined using FE-SEM. The STEM images of placebo PCL NPs-0 and NPs-4 are presented in Fig. 3 furthermore, the FE-SEM images are displayed in Fig. 4.
Uniform spherical NPs were observed by STEM, while near-spherical, irregular shaped NPs were observed by FE-SEM imaging, as observed in previous studies [18]. Partial agglomeration was observed in FE-SEM imaging, which can be attributed to the surface tension of water that affects the NPs during the drying process [25]. Additionally, the bridges observed between NPs in the FE-SEM images are formed due to the adhesive nature of the PVA used in the formulation. Removing PVA completely, even after particle washing, is challenging due to its inherently sticky nature [26]. FE-SEM images confirmed the formation of solid NPs with a purely polymeric structure. Overall, the morphologies of STEM images of NPs in liquid medium and the FE-SEM images of dried particles confirmed the manufacture of spherical and near spherical PCL encapsulated CBD NPs with O/W emulsion technique using ultrasonication.
3.3. Reaction Efficiency, Loading Capacity and Encapsulation Efficiency
The reaction efficiencies for placebo NPs and CBD-loaded PCL NPs were calculated to be 79.75% and 75.26%, 71.71%, 73.71%, and 70.91%, respectively, using Eq. 2 (Table 4). These results indicate that the reaction efficiency of placebo NPs is higher than that of CBD-PCL NPs. This means that the increased mass of hydrophobic bioactive ingredient in the reaction pot, decreases the number of manufactured NPs. Although the reaction experienced a 9% decrease in efficiency, this is still acceptable in terms of the function of the final product. The loading capacity of nanoparticles increased as the loaded ingredient amount increased. For instance, for the NP-4 formulation the loading capacity was about 22% with the starting 30mg CBD amount. This means that the 22% of NP-4 is composed of CBD by weight. The NPs seem to have capacity to be loaded by more CBD than used in this work.
The values of %EE for PCL NPs containing different amounts of CBD are given in Table 4. These results indicate a high loading of CBD into the PCL polymer matrix. The increase in the initial agent-to-polymer ratio resulted in an increase in loading, leading to encapsulation efficiencies ranging from 74.39–94.75%. This can be attributed to the high lipophilic nature of cannabidiol, which tends to be entrapped within the polymer matrix rather than dissolving in the external aqueous phase. This situation was also experienced by Pandey et al. that is, increased amount of betamethasone valerate per polymer amount resulted in increased encapsulation efficiency [27].
Table 4
Reaction efficiencies (n = 1) and encapsulation efficiencies of NPs (n = 3)
NPs | % Loading Capacity | % Reaction Efficiency | %Encapsulation Efficiency |
PCL NPs-0 | - | 79.75 | - |
PCL NPs-1 | 3.23 ± 0.03 | 75.26 | 74.39 ± 1.44 |
PCL NPs-2 | 6.87 ± 0.05 | 71.71 | 82.06 ± 0.19 |
PCL NPs-3 | 10.97 ± 0.08 | 73.71 | 88.74 ± 0.95 |
PCL NPs-4 | 22.05 ± 0.06 | 70.91 | 94.75 ± 0.61 |
3.4. CBD-PCL Solubility and Compatibility
Pure CBD, pure PCL, and PCL-CBD physical mixtures were evaluated in terms of their form and degradation profiles. The DSC thermograms of pure CBD, pure PCL, and PCL-CBD physical mixtures are demonstrated in Fig. 5, and the results are presented in Table 5. Consistent with previous studies, a sharp endothermic peak was observed at 69.82°C (corresponding to the melting point) for pure CBD, with an enthalpy of 77.45 J/g [26, 28]. Similarly, a sharp endothermic peak was observed at 68.81°C (corresponding to the melting point) for pure PCL, with an enthalpy of 77.45 J/g [11, 29]. This sharp peak is in accordance with the crystalline nature of CBD. The thermograms of PCL-CBD mixtures exhibited endothermic behavior corresponding to the melting endotherms of PCL polymer. For the CBD/ PCL ratios of 5/150, 10/150, 15/150, and 30/150 (mg/mg), the thermograms exhibited endothermic peaks at 68.93°C (enthalpy of 89.34 J/g), 68.81°C (enthalpy of 87.76 J/g), 68.71°C (enthalpy of 85.21 J/g), and 68.14°C (enthalpy of 82.46 J/g), respectively. In this work according to the thermal studies CBD powder, PCL powder and PCL-CBD mixtures exhibited a simple eutectic phase diagram. Accordingly, it may be concluded that physical mixtures of PCL and CBD have similar melting temperatures which may be implied that two components do not chemically interact with each other during melting.
Table 5
DSC analysis results of pure CBD, pure PCL and PCL-CBD physical mixtures
Materials | T(°C) | Enthalpy (J/g) |
Pure PCL | 68.81 | 77.45 |
Pure CBD | 69.82 | 77.45 |
5/150 | 68.93 | 89.34 |
10/150 | 68.81 | 87.76 |
15/150 | 68.71 | 85.21 |
30/150 | 68.14 | 82.46 |
3.5. In Vitro Release of CBD from PCL-NPs
The release kinetics of nanoparticles offer crucial insights into their capacity to influence drug release. Therefore, it stands as a significant parameter to be taken into account when evaluating the safety, effectiveness and quality of nanoparticles. In vitro drug release studies were conducted in buffer solutions at two different conditions, pH 1.2 and pH 7.4, to simulate gastric and intestinal conditions, respectively. The studies were carried out at 37°C for 8 h at pH 1.2 and for 96 h at pH 7.4. The aim was to develop a sustained dosage form suitable for the use of CBD in dietary supplements, due to the fact that the dietary supplement may remain in the stomach for a maximum of 8 h, in vitro release kinetics measurements were conducted at gastric pH 1.2 for 8 h to observe any stationary phase in drug release profile [30]. The release profile did not demonstrate the stationary release phase until 8 h release measurements.
The cumulative CBD release percentage at pH 1.2 was found to be significantly lower when compared to the CBD release at pH 7.4 from CBD-loaded PCL NPs, in 8 h. This indicates that the CBD-PCL NPs are pH-responsive nanocarriers for hydrophobic drugs as also found by Saqib et al. with their Amphotericin B loaded PCL NPs [31]. The slow and sustained release observed in 96 h may be attributed to the diffusion of CBD within the hydrophobic core of the PCL matrix [11, 32]. and is inherent to the hydrophobic nature of PCL. Crystalline and hydrophobic polymers generally exhibit slower degradation rates. Additionally, this slow release indicates that a small amount of CBD is adsorbed on the surface of PCL NPs. Snehalatha et al. reported that the degradation rate from PCL nanoparticle matrix is redated due to its hydrophobic nature and it was also reported that release of hydrophobic ingredients is also slower from hydrophobic matrices [33]. Wang et al. manufactured CBD encapsulated zein-whey protein composite nanoparticles. They observed about 20% CBD release at gastric fluid in 2h and then 90% of CBD was released at intestinal fluid after two additional hours [34]. Here the results imply that CBD/PCL nanoparticles allowed a limited CBD release when compared to the aforementioned study. As CBD being hydrophobic in nature, it exhibits limited release from the hydrophobic PCL nanoparticle matrix. The CBD/PCL combination exhibited the sustained release profile in this study.
A near-complete release of was observed for PCL NPs-4 (94.28%) at pH 7.4 at the end of 96 h. This limitation may be due to the usage of lower molecular weight of the polymer used in this study. It was previously reported that release from higher molecular weight polymers can be extended for months due to the slow degradation of the polymers in the release medium [35]. The release profile of CBD-loaded PCL NPs was fitted to zero-order, first-order, Higuchi, Korsmeyer-Peppas, and Hixson-Crowell kinetic models to evaluate the release kinetics best fitting to the CBD loaded NPs. Zero-order kinetic model refers to drug delivery systems that release a constant amount of active substance at the same time intervals, regardless of the concentration of the active substance [36]. On the other hand, the first-order kinetic model describes drug delivery systems where the amount of released active substance depends on the concentration of the active substance [37]. The Higuchi kinetic model explains the release of the active substance from homogeneous or heterogeneous matrix systems, while the Hixson-Crowell kinetic model illustrates the release of the active substance associated with changes in particle surface area [36, 38]. The Korsmeyer-Peppas kinetic model represents the release process where the diffusion and degradation of the polymeric structure are involved [39].
The correlation coefficients for the kinetic models of this work are presented in Table 6 and Table 7. The kinetic models with the highest correlation coefficients were depicted for the zero-order kinetic model at pH 1.2 and for the Higuchi kinetic model at pH 7.4.
In oral drug administration, the gastric retention time of the drug varies between 5 min and 5 h, depending on the individual's digestive state [40]. In the fasting state, this time can extend up to 8 h [30]. According to the results, the highest correlation coefficient for CBD-loaded NPs was observed in the Zero-Order Release Kinetic model. This model demonstrates that the NPs released the same amount of CBD at the same time intervals independent of its concentration [36]. This may indicate the CBD's partial presence on the particle surface.
Release kinetics of CBD from PCL NPs at pH 7.4 was found to fit to the Higuchi kinetic model. This model suggests that drug release from a homogeneous matrix system is controlled by the square root of time and occurs through Fickian diffusion [38]. Higuchi kinetics model presents a decrease in the amount of drug released over time. This is due to the initial release of the drug from the surface area, which forms the shortest diffusion path. However, as the dissolution process of the drug progresses, the surface area exposed to the dissolution medium decreases. After the depletion of the drug on the surface, the next layer of the film starts releasing the drug, leading to an increase in the length of the diffusion path. Additionally, the Higuchi model takes into account the relaxation of polymer chains that affect the drug release mechanism, in addition to Fickian diffusion, thereby demonstrating its impact on the drug release mechanism [41]. Furthermore, the results obtained at intestinal medium are consistent with other studies in the literature [41–43]. For instance, Kolluru et al. declared that hydrophobic Docetaxel encapsulated PCL NP system fitted to the Higuchi model at pH 7.4 [42]. Wadhwa et al. proposed that the molecular weight of the PCL polymer is one of the important factors determining the release rate from PCL nanoparticles [44].
Table 6
Correlation coefficients of the release kinetic models at pH 1.2
NP | Zero Order | First Order | Korsmeyer-Peppas | Higuchi | Hixson-Crowell |
PCL NPs-1 | 0.9878 | 0.9846 | 0.9883 | 0.8739 | 0.9859 |
PCL NPs-2 | 0.9813 | 0.9760 | 0.9839 | 0.8509 | 0.9781 |
PCL NPs-3 | 0.9843 | 0.9790 | 0.9804 | 0.8601 | 0.9811 |
PCL NPs-4 | 0.9956 | 0.9957 | 0.8222 | 0.8739 | 0.9959 |
Average | 0.9873 | 0.9838 | 0.9687 | 0.8647 | 0.9853 |
Table 7
Correlation coefficients of the release kinetic models at pH 7.4
NP | Zero Order | First Order | Korsmeyer-Peppas | Higuchi | Hixson-Crowell |
PCL NPs-1 | 0.9619 | 0.9908 | 0.9541 | 0.9883 | 0.9868 |
PCL NPs-2 | 0.9476 | 0.9882 | 0.9035 | 0.922 | 0.9838 |
PCL NPs-3 | 0.9659 | 0.9783 | 0.9330 | 0.9907 | 0.9909 |
PCL NPs-4 | 0.9414 | 0.9621 | 0.8569 | 0.9883 | 0.9852 |
Average | 0.9542 | 0.9799 | 0.9119 | 0.9899 | 0.9867 |
3.6. Internalization of PCL NPs into the Cells
Internalization of Nile Red encapsulated PCL NPs into L929 fibroblast cells was visualized under a fluorescent microscope (Fig. 7). It is apparently seen that PCL NPs were internalized into the cells and the dye was carried into the cytoplasm. This shows the PCL NPs encapsulation system may also be used as a sustained drug delivery system to carry bioactive agents into the cells. It was also previously reported that internalization of PCL NPs into the cells may be caused by random interaction of the cell membrane and polymer [42].
3.7. In vitro cell proliferation studies
An in vitro cell proliferation assay was conducted to determine the effects of the active compound CBD on L929 normal fibroblast and MCF-7 breast cancer cells. According to the test results, MCF-7 cells, which are used as a model for breast cancer, exhibited a resistant proliferation profile compared to L929 cells in response to CBD. The IC50 values calculated at different time intervals are listed in Table 8.
Table 8
The effect of CBD active compound on cell proliferation
Time (h) | IC50 (µM) |
| Cell line |
| L929 | MCF-7 |
24 | 0.79 ± 0.08 | 19.88 ± 0.87 |
48 | 1.12 ± 0.05 | 18.73 ± 0.75 |
72 | 6.17 ± 0.12 | 18.17 ± 0.82 |
The antiproliferative effect of CBD on L929 cells decreased as the duration of cell exposure increased. The cytotoxic effect of CBD on normal fibroblast cells decreased at 24, 48, and 72 h, respectively (Table 8). This could be attributed to the decrease in CBD concentration per cell as the cell number increased. The antiproliferative effect of CBD on MCF-7 breast cancer cells, although less than that on L929 cells, did not show a significant difference over time. CBD had a 3-fold greater antiproliferative effect on normal fibroblast cells after 72 h compared to its effect on breast cancer cells. This may be due to the drug resistance of cancer cells to various active substances. According to the results of our Molecular Docking study recently published, the LD50 value of CBD when administered intravenously in rats was 799 mg per kg and was found to be safe in oral administration [45]. In their study, Mato et al. reported that the application of 10 µM CBD for 24 h resulted in a 30% antiproliferative effect on oligodendrocyte nerve cells [46]. The results here indicate that CBD has a reducing effect on the proliferation of both normal and cancer cells (Fig. 8, Fig. 9). In order to evaluate the effect of CBD in cancer treatment, nano formulations may be developed to target cancerous tissues in order to avoid damaging normal cells. In the design of the NP formulation targeting cancer cells, specific modifications targeting cancer tissues will be required. These modifications should enable the anticancer CBD-NP formulation to reach the cancer tissue without damaging normal tissues.