Polymeric SL-loaded NCs were fabricated using the nanoprecipitation method, and the impact of formulation variables, such as PEG density on the PLGA chain and the type of oil core, on the physicochemical characteristics of the SL-loaded NCs was investigated. Mean particle size, PDI, zeta potential, encapsulation efficiency, in vitro release profile and in vitro stability were determined. The aim of the work was to use this data and select an optimised formulation for further in vivo evaluation. This optimised formulation was considered to be one with the smallest diameter, best encapsulation efficiency, longest shelf-life stability and controlled therapeutic action.
Effect of polymer type on physicochemical properties
Modification of PLGA by incorporation of PEG blocks within the polymeric structure is a common experimental variable within the formulation strategy of a nanoparticulate system. Effects on aqueous solubility of poorly soluble payloads, aggregation, stability and opsonisation are often reported [12, 23]. However, the approach is also associated with further advantages, such as improvement in antitumour efficiency and targeting characteristics [19, 20]. Therefore, the physicochemical properties of different SL-loaded NC (F3, F6 and F9) fabricated from three different PLGA polymers were investigated (Fig. 1). PLGA-based NC (F3) (196.5 ± 20.9 nm) were significantly larger in size (p < 0.01) than NC prepared from 10% PEGylated polymers of (F9) (117.5 ± 11.5 nm). A non-significant difference was observed when compared to 5% PEGylated NC (F6) (177.4 ± 14.5 nm) (p ˃ 0.05). SL-loaded NC prepared from 10% PEG-PLGA diblock copolymers (F9) were significantly smaller than NC prepared from 5% PEG-PLGA copolymer (p < 0.01), (Fig. 1a). Similar results were observed during the preparation of polymeric nanospheres using the same polymers [22, 23]. This was attributed to the effect of linked PEG molecules on the assembly of smaller NC. Modification of PLGA backbone by addition of PEG as a hydrophilic polar segment can affect the physicochemical properties of the SL-loaded NCs prepared from PEGylated copolymers. The particle size is often increased by an increase in the hydrophobic segment of the used copolymer. Higher PEG content resulted in a smaller NP size, attributable to the short chain length of 10% PEG-PLGA copolymer compared to 5% PEG-PLGA and PLGA [23, 27]. All SL-loaded NC showed low polydispersity index ranging from 0.19 to 0.38. Increasing the PEG content in the NC polymeric coat resulted in a significant increase in PDI values compared to PLGA NC (p < 0.05).
The ζ-potential of NC colloidal suspension plays an important role in maintaining in vitro stability. In vitro stability of colloidal suspension depends on the degree of electrostatic repulsion between nanoparticles. Higher negative or positive zeta potential values resulted in good in vitro stability with low tendency of particles aggregation . F3 exhibited higher negative ζ-potential values (− 21.9 ± 2.8 mV) compared to the PEGylated PLGA NC (F6 and F9) (p < 0.001) (Fig. 1b). Increasing the PEG content in the backbone of the polymeric coat from 5 to 10% had a non-significant effect on the ζ-potential (p ˃ 0.05). The lower negative ζ-potential of PEGylated NC is due to the presence of surface-decorated PEG chains which mask the free negative surface charge represented by the free anionic groups of PLGA [22, 23].
The presence of PEG blocks in the copolymer structure resulted in a significant increase in SL entrapment (p < 0.001). Encapsulation efficiency was increased from 68.8 ± 6 in PLGA to 92.18% ± 1.8% in 10% PEG-PLGA (Fig. 1c). This can be explained by the amphiphilic nature of both the polymer itself and SL. PEG regions of the copolymer will most likely orientate towards the continuous aqueous phase, which can associate with the hydrophilic regions of the SL molecule. The hydrophobic part of the SL molecule will orientate towards the oil core. Therefore, the PEG-PLGA diblock copolymer creates a suitable microenvironment created by its amphiphilic nature, leading to high SL loading .
The in vitro release profile showed comparable burst release of SL from the three different polymers (p ˃ 0.05). Burst release after 24 h was 53.2% ± 8.9, 60% ± 4.2 and 50% ± 8.3 from F3, F6 and F9, respectively (Fig. 1d). The observed burst release can be attributed to the high amount of SL attached to the surface of PLGA and PEGylated due to its amphiphilic nature and surface-active properties. Release of SL from PLGA NC (F3) was sustained after the first 24 h, and release became slower, with approximately 70% of SL released over the experimental period of 72 h. The PEGylated PLGA NC showed the faster release of SL after the first 24 h with approximately 90% and 97% of SL released over 72 h from F6 and F9, respectively.
Effect of oil core type
SL was poorly soluble in palm oil (1.06 mg ml−1), being significantly lower than that in both castor oil (1.79 mg ml−1) and glyceryl monocaprate (2.21 mg ml−1). The physicochemical characteristics of different SL-loaded NC fabricated using these different oil cores are shown in Fig. 2. SL-loaded NC based on glyceryl monocaprate (F9) exhibited significantly lower size (117.5 ± 11.5 nm) compared to NC-palm oil (F8) (162.5 ± 17.2 nm) and NC-castor oil (F7) (179.2 ± 13.1 nm) for the 10% PEGylated copolymer types (p < 0.05). However, this trend was not apparent in the PLGA NC (Fig. 2a). Type of oily core had a significant impact on the physicochemical properties of SL-loaded NCs. At room temperature, vegetable oils such as palm oil and castor oil showed higher viscosity relative to glyceryl monocaprate. The smaller size (p < 0.05) observed for (F9) is probably a consequence of the lower viscosity of the glyceryl monocaprate compared with the palm oil and castor oil used for preparation of F8 and F7, respectively . The PDI values (data not shown) for glyceryl monocaprate cores were significantly lower than the palm oil and castor oil formulations. Variation in oil type did not have any significant effect on the ζ-potential of the different NC formulations (p ˃ 0.05) (Fig. 2b).
The % encapsulation efficiency of all SL-loaded NC ranged from 56.3 ± 5.9 to 92.18% ± 1.8%. Glyceryl monocaprate-based NC exhibited the highest encapsulation efficiency values compared to other oil-based formulation, regardless the type of the polymer (Fig. 2c). This is attributed to solubility of SL in glyceryl monocaprate, which prevented its undesirable loss towards the aqueous phase leading to a high degree of entrapment. In addition, glyceryl monocaprate has a lower hydrophilic–lipophilic balance (HLB approximately 5–6) than castor oil (HLB 14) and palm oil (HLB 10). Therefore, NC based on a glyceryl monocaprate core may offer a favourable microenvironment for encapsulating water-insoluble glycolipids, such as lactonic SL . This finding has been demonstrated in other work, where the highest encapsulation efficiency of paclitaxel was observed in glyceryl monocaprate nanocapsule cores compared with that in other oils (palm oil and coconut oil) . Similar findings are observed for entrapment of curcumin into castor oil-based nanocapsule due to enhanced solubility in the core medium .
In vitro release profiles showed that changing the oil core influenced the initial burst release from different NC (Fig. 2d). In PLGA-based NC, a higher burst release had been observed from F3 (containing glyceryl monocaprate) compared to F1 and F2 containing castor oil or palm oil, respectively. The opposite effect was then observed for PEGylated NC under the same conditions. The initial burst release after 24 h was less for glyceryl monocaprate compared to other oils. This might be attributed to a polymer-oil interaction driven by the relative hydrophilicity of the polymer. The higher burst release for PLGA NC might be due to its favourable association with glyceryl monocaprate, which led to longer solidification time and, consequently, to the formation of more porous NC structure, facilitating drug release .
Transmission electron microscopy
The appropriate morphology of nanoparticulate drug delivery system is crucial for the successful design of therapeutic drug delivery systems. In this work, TEM was used to visualise the particulate and to evaluate aggregation or adhesion of SL-loaded NC. TEM images revealed a smooth spherical shape of (F9) homogenous size with no aggregations (Fig. 3). The average size obtained from TEM was comparable to what obtained by light scattering.
In vitro stability of SL-loaded NC
After formulation of all SL-loaded NC, non-entrapped drug was removed by gel filtration chromatography. The purified NC formulations (F3, F6 and F9) were selected for further study as they showed superior physicochemical properties for each polymer type compared to other NC. These formulations were characterised for their size, PDI, ζ-potential immediately after purification at time zero and every 10 days over 1 month of storage at 4 °C (Table 2). The results from assessment of stability parameters, such as time-dependent variation in particle size, confirmed that F3 and F9 were sufficiently stable for 1 month after preparation. There was a non-significant (p ˃ 0.05) increase of size, PDI and zeta potential compared to freshly prepared ones. Although F6 prepared from 5% PEG-PLGA diblock copolymer showed non-significant (p ˃ 0.05) increase in size and zeta potential, it showed a significant (p ˂ 0.05) increase in PDI value after 30 days from preparation. No apparent change in colour, clarity or phase separation was observed in any of NC preparations. Herein, stability results showed that SL-loaded NCs prepared from 10% PEG-PLGA copolymer were biocompatible, stable and monodisperse nanosystem for SL encapsulation, with potential applications for drug delivery. Our results prove that nanoprecipitation technique leads to efficient stabilisation of oily core PEGylated NCs. The in vitro stability measurements showed sufficient colloidal stability for 1 month. Stability results confirmed lack of evident destabilisation phenomena associated with no degradation over time which can affect the size, PDI and zeta potential of the prepared NCs.
Cellular uptake of SL-loaded NC
This study investigated the efficiency of cellular uptake of SL-loaded PLGA and PEGylated PLGA NC by CT26 colon cancer cells. The effect of polymer type on the in vitro cellular uptake results are shown in Figs. 4 and 5. The quantitative flow cytometry analysis was used to detect the uptake of the coumarin 6-tagged SL-loaded NC (F3, F6 and F9) in CT26 colon cancer cells after 24 h of treatment (Fig. 4a–d). The intracellular uptake of fluorescent NC into CT26 cells was expressed by a fluorescence shift from control cells depending on the fluorescence intensity (Fig. 4a–c). The extent of cellular uptake was measured secondly by the percentage of cells with positive staining following treatment (Fig. 4d). After 24 h, F9 exhibited significantly higher cellular uptake (p < 0.001) than other NC formulations with lesser amounts of PEG in the polymer coat (F3 and F6). Interestingly, these results were in good agreement with the optimum physio-chemical properties of F9, with the lowest particle size, highest PEG content and lowest zeta potential . These parameters gave rise to the greatest uptake in CT26 cells. Previous results supported the higher cellular uptake of PEGylated PLGA NC in CT26 cells, due to PEG pronounced effect on decreasing nanoparticles size and minimising the electrostatic repulsion force with the cell membrane .
Given that SL and the other formulation components possess no fluorescent activity, it was necessary to add a fluorescent dye to determine the position of NC following exposure in the test cell lines. Images from fluorescence microscopy supported particulate uptake in CT26 cells (Fig. 5). The SL-loaded NC (F9) were primarily localised in the cytoplasm, while some fluorescence intensity was observed around the nucleus, which supports possible intracellular localisation of NC (Fig. 5g–i). Cells treated with coumarin 6 solution showed minimal uptake as the dye was localised at the outer cell surface (Fig. 5d–f). Based on these quantitative and qualitative cellular uptake results, it can be concluded that SL-loaded 10% PEGylated PLGA NC (F9) showed maximum intracellular uptake.
In vitro cytotoxicity
An assessment of the physicochemical data for all formulations used in this work allowed selection of F9 as an optimised formulation. This was based on F9 having the smallest size, highest %EE, optimum shelf-life stability and highest cellular uptake. Therefore, F9 was selected for further cell viability and organ distribution studies.
The cytotoxic action of SL-loaded NC (F9) on CT26 colon cancer cells and CCD-841-CoN normal colon epithelial cells was evaluated by MTT assay after 24, 48 and 72 h of treatment with 20, 40 and 60 μM of free SL and SL-loaded NC (Fig. 6a). Blank NC had no cytotoxicity action on colon cancer cells. The dose-response curves were generated to detect the IC50 which is the drug concentration resulting in 50% growth inhibition of colon cancer cells. Free SL showed a dose-dependent effect as cell viability decreased with increasing dose from 20 to 60 μM. Significant reduction in cell viability compared to control (p < 0.05) was observed. Treatment with SL-loaded NC (F9) reduced cell viability in a dose-dependent effect meanwhile it maintained its cytotoxic action up to 3 days after treatment due to sustained drug release. The mean IC50 value for SL-loaded NC (F9) in CT26 cells was approximately 60 μM, which was achieved within 24 h. Different doses of free SL showed a slight decrease in cell viability of CCD-841-CoN normal colon epithelial cells especially after 3 days of treatment (Fig. 6b). Twenty-five percent reduction of cell viability was observed after treatment with the highest dose of 60 μM of free SL after 72 h. On the other hand, SL-loaded NC (F9) achieved non-significant (p ˃ 0.05) cytotoxic effect on normal colon epithelial cells within the concentration range used in this study. This confirms the safety profile of the designed drug delivery system compared to free drug for further in vivo applications. Interestingly, these results are different from other studies . This might be due to the variation of purity of SL used, due to different sources also due to the lower range of drug concentration used in this study. These results showed that NC drug delivery system can deliver SL to its subcellular site and protect it from possible degradation to achieve lower cell viability and greater cytotoxicity when compared to the free SL after 24, 48, and 72 h of treatment (p < 0.001).
In vivo tumour uptake and organ biodistribution study
Tumour accumulation and organ biodistribution of DiR-tagged SL-loaded NC (F3 and F9) were assessed in vivo in mice bearing the murine colon cancer CT26 model. Whole-body in vivo imaging was performed at 1, 4 and 24 h after injection (Fig. 7). Imaging of excised organs, such as the heart, lung, liver, spleen, kidneys and the tumour mass is shown in Fig. 8.
In vivo imaging data from mice treated with PBS (control group) confirmed that no background fluorescence was present (data not shown). SL-loaded NC (F3) formulated from PLGA polymer showed no tumour uptake after 1, 4 and 24 h, but did confirm accumulation in the liver and spleen (Fig. 7). Tumour uptake of 10% PEGylated NC (F9) was observed in both implanted tumours when imaged from the dorsal and ventral view.
The excised organs from control animals (treated with PBS) showed no fluorescence. Animal treated with F3 showed high fluorescence from both liver and spleen, but low fluorescence from the tumour masses. Excised organs from animals treated with F9 showed the highest fluorescence intensities from the tumours when compared to other organs (Fig. 8a). Fluorescence intensities from each organ were further quantified (Fig. 8b) showing mean fluorescence intensities from tumours treated with F9 to be highly significant compared to animals treated with F3 (p < 0.001) which confirms high tumour accumulation of F9. Fluorescence intensity quantified from the liver in animals treated with F3 was highly significant compared to animals treated with F9 (p < 0.001).
Size and surface properties of nanocarriers play an important role in controlling drug pharmacokinetics, biodistribution and drug resistance in some cancer cells [32, 33]. Nanoparticles ≥ 200 nm and without surface modification will induce the host immune systems to be engulfed by hepatic Kupffer cells and other reticuloendothelial cells, resulting in their rapid clearance from the circulation system. Nanoparticles with size ≤ 150 nm and PEG modification appear to be appropriate for passive targeting as they are better able to exploit site-specific endothelial irregularities and histological variability [28, 34, 35]. An appreciation of this often-debated EPR effect in solid tumours is one of the essential prerequisites for designing a successful antitumour drug delivery. As shown in this study, F9 fulfils the requirements for a long-circulating NC as its size (117.5 ± 11.5 nm) and PE-decorated surface are considered to be desirable features. F9 is, therefore, optimum for passive targeting to colon cancer due to its preferable biodistribution, which will lead to the highest therapeutic efficacy in vivo.
In vivo antitumour activity
A low number of studies have investigated the therapeutic potential of SL in vivo. A sepsis model has been used, which showed that SL mixtures containing both acidic and lactonic SL reduced mortality in rats with experimentally induced sepsis via cecum puncture . Recently, in vivo anticancer models have been used to study the anticancer activity of SL against human cervical cancer .
All animal involved in this study appeared healthy throughout the whole experiment duration and no significant (p ˃ 0.05) loss in body weight was detected. There were no toxicity signs after treatment with either free SL or SL-loaded NC (F9) at the drug concentration used. The dose used for this study was 10 mg SL per kg of mice weight, and 5 injections were administrated in vivo to the mice. The in vivo tumour growth inhibition study in mice bearing CT26 colon cancer after SL treatment showed no antitumour activity of blank NC and a pronounced anticancer effect of SL-loaded NC (F9) (Fig. 9a). The average tumour size of animals treated with F9 was significantly lower compared to the control group and animals treated with the free drug (p < 0.001). At the end of the experiment, the average size of the tumour mass of control animals treated with saline was found to be 993.5 ± 142.5 mm3. The average tumour size of animals treated with free SL was 706.7 ± 105.3 mm3 and the percentage tumour growth inhibition (% TGI) was 28.9%. A more pronounced effect was recorded in animals treated with F9. There was a significant reduction in the volume of tumour mass at each time point starting from the 10th day till the end of the experiment when compared with control and free drug-treated group. The final average size of tumour was found to be decreased to 426.6 ± 115.4 mm3 and the % TGI was 57.09% (Fig. 9b). It was observed that the F9 achieved a significant reduction in tumour size (p < 0.001) confirming its greater antitumour activity than free SL.
At the end of the in vivo experiment, all mice were sacrificed, and implanted tumours were excised and weighed (Fig. 9c). The change in tumour weights was observed in the case of animals treated with free SL or F9. Treatment of mice with F9 reduced the average tumour weight to 0.62 ± 0.1 g compared to 1.6 ± 0.46 g for control animals with a percentage reduction of 61.25%, which is highly significant compared to control (p < 0.001) and free drug-treated mice (p < 0.01). The percentage reduction in tumour weights was 37.5% in mice treated with free SL, which is statically significant compared to control mice (p < 0.01).