Curcumin (CUR), a polyphenol, is the main ingredient extracted from the dried rhizome of Curcuma longa L. (turmeric) and has various pharmacological effects including antioxidant, anti-inflammatory, antibacterial, immunomodulatory and anti-ischemic activity (Afolayan et al. 2018; Kannigadu and N’Da 2021; Lin et al. 2022; Marton et al. 2022). Furthermore, CUR is a promising treatment for colon, pancreas, breast, and lung cancer through the inhibition of cancer proliferation, metastasis, and angiogenesis (Devassy et al. 2015; Mirzaei et al. 2016). CUR has been found to inhibit tumor growth through many cell signaling pathways including cell proliferation pathways (i.e., cyclin D1 and c-myc), cell survival pathways (i.e., Bcl-2, Bcl-xL, cFLIP, XIAP, and cIAP1), caspase activation pathways (i.e., caspase-8, -3 and -9), tumor suppression pathways (i.e., p53 and p21), death receptor pathways (i.e., DR4 and DR5), and protein kinase pathways (c-Jun N-terminal kinases [JNK], protein kinase B, and 5’-adenosine monophosphate-activated protein kinase) (Huang et al. 2017; Kocaadam and Şanlier 2017; Zhang et al. 2018; Li et al. 2022). Moreover, in recent years, evidence has emerged that CUR affects several molecular and cellular pathways involved in melanoma development, such as MST1, JNK, Foxo3, Bim-1, Mcl-1, BCl-2, Bax and JAK-2/STAT-3 (Zhang et al. 2015; Mirzaei et al. 2016). Melanoma is one of the most aggressive types of skin cancer, with high mortality. Its incidence has increased at a faster rate than any other type of cancer since the mid-1950s (Erdei and Torres 2010). Most melanomas can be treated surgically in the early stages, but they have a high metastatic potential. Due to the lack of effective treatment and resistance to existing chemotherapy in the late stages of melanoma, the average survival time is only 3-11 months and the 5-year survival rate is less than 5% (Mirzaei et al. 2016).

Although CUR has demonstrated therapeutic effects on several cancer cells including melanoma, colon, pancreas, breast, and lung cancers, its use is limited due to its low solubility, P-glycoprotein (P-gp) involvement, and rapid metabolism (Shan et al. 2022). Orally administered CUR is excreted into feces in unabsorbed forms (75% of dose) and 60% of absorbed CUR is pumped out via P-gp or metabolized in the intestine (Ravindranath and Chandrasekhara, 1980). Overall, 45% of the CUR clinical trials have aimed to enhance bioavailability and improve the pharmacokinetic properties of CUR (Panknin et al. 2023). Many researchers sought to overcome these limitations through the use of surfactants and drug delivery systems, such as phospholipids, polymer nanoparticles, metal or non-metal nanoparticles, liposomes, and self-emulsifying drug delivery systems (He et al. 2010; Bollu et al. 2016; Briskey et al. 2019). Among these technologies, mesoporous silica nanoparticles are one of the most promising technologies. The superiority of these nanoparticles is due to (i) variable and controllable size, (ii) stability from heat, pH, and chemical degradation, (iii) high surface area and large pore volume, (iv) easy attachment of various ligands, metal composites, biomaterials to external and internal surfaces (Bollu et al. 2016; Kotcherlakota et al. 2016). For this reason, many researchers have attempted to improve drug bioavailability and anti-cancer activity using mesoporous silica nanoparticles. Su et al. manufactured a complex particle by adding CUR to SBA-15, a mesoporous substance, and demonstrated its therapeutic effect on metastatic lung cancer (Su et al. 2019). Moreover, two types of mesoporous silica nanoparticles (i.e., MSU-2 and MCM-41) that contain CUR have shown better bioavailability and greater anti-cancer activity than CUR itself (Bollu et al. 2016).

Various surfactants that have the potential to modulate drug metabolism enzymes and efflux transporters are used in conjunction with drug delivery systems to increase aqueous solubility and enhance drug bioavailability (Williams et al. 2013; Choi et al. 2023). For example, a poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) triblock copolymer, such as pluronic P85 (PP85), can improve the stability and solubility of drugs by forming a core (Kabanov et al. 2002; Oh et al. 2004; Samanta and Roccatano 2013). Moreover, it has been found to accumulate P-gp substrate chemotherapeutic drugs in tumor cells, thereby increasing the efficiency of anti-cancer effects (Kwon et al. 2020). Therefore, we aimed to formulate CUR- and PP85-loaded mesoporous nanosilicates. We also aimed to characterize the formulation and evaluate the pharmacokinetic properties and the anti-cancer activity of this CUR- and PP85-loaded nanosilicate formulation in B16F10 melanoma-bearing mice.

Materials

CUR, PP85, D-α-Tocopherol polyethylene glycol 1000 succinate (TPGS), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), sodium chloride, sodium bicarbonate, potassium chloride, potassium phosphate dibasic trihydrate, magnesium chloride, calcium chloride, sodium sulfate, tris (hydroxymethyl) aminomethane, dimethyl-sulfoxide were purchased from Sigma-Aldrich Chemical CO. (St. Louis, MO, USA). Fetal bovine serum, dulbecco’s modified eagle’s medium (DMEM), and penicillin-streptomycin were purchased from Hyclone Laboratories (Logan, UT, USA). All other reagents used were of reagent grade.

Mesoporous nanosilicate, named as Smart Mesoporous Ball 7 (SMB7; Patent no. KR-10-2240246), was obtained from CEN CO., Ltd (Miryang-si, Gyeongnam, Republic of Korea). It contains SiO2 and ZnO (atomic composition: Si 30.4%, O 68.0%, and Zn 1.6%). The average particle size of SMB 7 was 250 nm, with pore diameter and volume of 2.54 nm and 0.81 cm³/g, respectively.

Preparation of CUR-loaded mesoporous nanosilicate and fabrication with PP85

To select the surfactant that has P-gp inhibitory effect, we measured the P-gp reversal efficacy of Cyclosporin A (CsA), TPGS, PP85, and SMB7 in P-gp overexpressed cells (i.e., LLC-PK1-P-gp cells), according to the method of Kwon et al. (Kwon et al. 2020). Briefly, CUR (5 μM) was added to 96 well plate seeded LLC-PK1-P-gp cells at a density of 5×10⁴ cells/well in the presence or absence of CsA (20 μM) or varying concentrations of PP85, TPGS, and SMB7 (1, 10, 100, and 100 μM) and incubated for 30 min in CO₂ incubator. Cells were, then, washed twice with 200 μL aliquot of ice-cold PBS. One hundred fifty microliters of berberine methanol solution (1 ng/mL, internal standard [IS]) were added to the cells and incubated for 20 min in 4°C cold room. Aliquots (100 μL each) of the cell aggregates centrifuged at 16,000 × g for 5 min. Supernatants (50 μL each) were transferred to the autosampler vials and 5 μL aliquots were injected into the liquid chromatography with tandem mass spectrometry (LC-MS/MS) system.

To optimize the ratio of PP85 to CUR, the aqueous solubility of CUR was measured with varying ratios of PP85. To optimize the ratio of CUR and PP85 to SMB7, CUR- and PP85- loaded mesoporous nanosilicates (CUR-PP85-SMB7) in the CUR: PP85: SMB7 ratio of 1:2:0 to 1:2:5 was prepared and the aqueous solubility of CUR from these formulations were measured.

CUR-loaded mesoporous nanosilicates (CUR-SMB7) formulation were prepared by solvent evaporation method. Briefly, 1 g of CUR was dissolved in 20 mL of methanol:acetone=1:1 (v/v) mixture in a round bottom flask, then, 4 g of SMB7 mesoporous nanosilicates were added into the CUR solution and stirred for 1 h. The solvent was evaporated using a vacuum evaporator.

To fabricate PP85 into CUR-loaded mesoporous nanosilicates, the freeze-drying method was applied. Briefly, 0.5 g of pluronic P85 and 2.5 g of CUR-loaded mesoporous nanosilicates were dispersed in 50 mL of distilled water and freezed at –80°C overnight. The freezed dispersion of 0.5 g of PP85 and 2.5 g of CUR-loaded mesoporous nanosilicates were connected to the freeze dryer (FDCF-12012, Operon, Gyeonggi-do, Republic of Korea) and lyophilized under vacuum conditions at −120°C for 72 h. The lyophilized PP85- and CUR-loaded mesoporous nanosilicate powder was then dried in a desiccator for 1 day. The solubility, dissolution profile, and formulation characteristics of the lyophilized PP85- and CUR-loaded mesoporous nanosilicate powder was evaluated, as previously described (Lim et al. 2022; Jeon et al. 2024).

Characterization of CUR-loaded mesoporous nanosilicate formulation

Drug loading (DL, %) and encapsulation efficacy (EE, %) of CUR in the mesoporous nanosilicate formulation (i.e., CUR-SMB7 with ratio of 1:4 (w/w) and CUR-PP85-SMB7 with ratio of 1:2:4 (w/w)) was measured. Five mg of the powder was placed in 50 mL of methanol and completely dissolved using a shaker (Thermo micro-mixer, confide-S20H, Gunpo-si, Gyeonggido, Republic of Korea) at 37°C for 2 h, and then, filtered through a 0.45 μm membrane filter. The filtrate was diluted 100-fold with mobile phase and injected into the LC-MS/MS system to measure the CUR concentration. DL and EE of CUR were calculated using the following equations (Jeon et al. 2024):

$$\text{DL(\%)=}\frac{\text{CUR content (mg) in formulation}}{\text{total weight of formulation (mg)}}\times 100\dots \dots \dots$$(1)

$$\text{EE(\%)=}\frac{\text{CUR content (mg) in formulation}}{\text{Weight of initial CUR(mg)}}\times 100\dots \dots \dots$$(2)

To evaluate the solubility of the formulation, 5 mg of the powder (CUR, CUR-SMB7, and CUR-PP85-SMB7) was weighed, and then, added in 5 mL distilled water and simulated body fluid (SBF, pH 7.4) consisting of 135 mM sodium chloride, 4.2 mM sodium bicarbonate, 4 mM potassium chloride, 1 mM potassium phosphate dibasic trihydrate, 1.5 mM magnesium chloride, 2.5 mM calcium chloride, 0.5 mM sodium sulfate 0.071 g, tris(hydroxymethyl)aminomethane 50 mM, and 40 mM hydrochloride (Kokubo et al. 1990). The mixtures were shaken at 37°C for 2 h. After 1 min of centrifugation, the supernatant was filtered through a 0.45 μm membrane filter. The filtrate was diluted 100-fold with mobile phase and injected into LC-MS/MS to measure the CUR concentration.

In vitro drug release of CUR from the formulation was measured as follows: CUR, CUR-SMB7, and CUR-PP85-SMB7 (all contain 2 mg CUR) were placed in the cellulose membrane pouch (Snakeskin^TM Dialysis Tubing 10,000 MWCO, Thermo scientific, USA) and 5 mL of SBF (pH 7.4) was added inside of the cellulose membrane pouch. The cellulose membrane pouch was put into 50 mL of SBF containing 0.5% tween 80 in a 250 mL beaker and shaken at 100 rpm in a dual-motion shaker (FINEPCR, Gunpo-si, Gyeonggi-do, Republic of Korea). Samples (100 μL aliquots) were collected at 0, 0.5, 1, 2, 4, 8, 12, 24, 36, 48, 60, and 72 h, and then, diluted 100-fold with mobile phase and injected into the LC-MS/MS system to measure the CUR concentration.

The particle size and surface image of CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 were observed under scanning electron microscope (SEM) (SU8220, Hitachi, Tokyo, Japan). The morphological change of SMB7, CUR-SMB7, and CUR-PP85-SMB7 during 48 h in SBF (pH 7.4) using a bio-transmission electron microscope (Bio-TEM). A drop of SMB7, CUR-SMB7, and CUR-PP85-SMB7 suspension was placed on a 200-mesh copper grid coated with carbon, and the samples were air dried. The samples were examined at 100 kV in Hitachi HT 7700 TEM (Hitachi High-Technologies Corporation, Tokyo, Japan).

The X-ray diffraction (XRD) scanning of the CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 was performed using an Empyrean X-ray diffractometer (Malvern Panalytical Ltd., Malvern, England) using Cu Kα radiation at 40 mA and 40 kV. Data were obtained from 5-70° with a step size of 0.02° and a scanning speed of 5°/min using 10 mg of a sample formulation. Differential scanning calorimetry (DSC) thermograms of the CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 were determined using a DSC Q2000 (TA Instruments, New Castle, DE, USA). Approximately 10 mg of a sample was placed in a closed aluminum pan and heated with a scanning rate of 5°C/min from 10 to 250°C, with nitrogen purging at 20 mL/min. The temperature scale was calibrated using indium. The Fourier-transform infrared spectroscopy (FT-IR) spectra of CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 were obtained in the spectral region of 4,000-600 cm⁻¹ with a resolution of 4 cm⁻¹ and 64 scans using a Frontier FT-IR spectrometer (PerkinElmer, Norwalk, CT, USA) in the transmittance mode.

Cytotoxicity of CUR on B16F10 cell

B16F10 cells were purchased from Korea Cell Line Bank (Seoul, Republic of Korea). B16F10 cells were grown in a petri dish containing DMEM supplemented with 10% fetal bovine serum and 1% penicillin streptomycin. The cells were maintained at 37°C in a humidified atmosphere with 5% CO₂/95% air. The cells were seeded in 96 well plates at a density of 10⁴ cells/well and incubated for 24 h.

The cytotoxicity of CUR, CUR-SMB7, and CUR-PP85-SMB7 were assessed in B16F10 cells and the incubation medium was replaced with a culture medium containing varying concentrations of CUR, CUR-SMB7, and CUR-PP85-SMB7 (CUR concentrations of 0, 0.1, 0.2, 0.5, 1, 10, 50, and 100 µM) for 48 h. The culture medium was changed every 24 h. The cytotoxicity of SMB7 was also assessed using the same protocol with the SMB concentrations ranging from 0.368 µg/mL to 368 µg/mL. After 48 h incubation, the culture medium was aspirated and 200 µL of the culture medium containing 0.5 mg/mL MTT was added to each well, and the plate was incubated for 4 h. The medium was aspirated, and the purple formazan product was dissolved in 120 µL of dimethyl sulfoxide.

The resultant cellular content was determined by measuring the absorbance at 590 nm, and the viability of the treated cells was expressed as a percentage of live cells compared with control group after background correction. Relevant data were fitted to an inhibitory effect model [i.e., $v=E_{\max }\left(1-\frac{[/]}{E{C}_{50}+[/]}\right)$]. The EC₅₀ value reflects the concentration of CUR or SMB7 to show half-maximal inhibition and [I] represents the concentration of CUR or SMB7.

Pharmacokinetic study of CUR

Male C57BL/6 mice (7 weeks, 23-25 g) were purchased from Samtako Co. (Osan, Republic of Korea) and acclimated for one week in an animal facility at Kyungpook National University. Food and water were provided ad libitum. All animal procedures were approved by the Animal Care and Use Committee of Kyungpook National University (Approval No. 2023-0576) and conducted according to the National Institutes of Health guidelines for the care and use of laboratory animals.

C57BL/6 mice were randomly divided into three groups (the CUR, CUR-SMB7, and CUR-PP85-SMB7 groups), with 8 mice in each group. The CUR, CUR-SMB7, or CUR-PP85-SMB7 were suspended in saline at a concentration of 5 mg CUR equivalent/3 mL and administered by intraperitoneal injections (5 mg/kg). At 0.25, 0.5, 1, 2, 4, 8, and 24 h after administration, approximately 100 µL of blood was collected from the orbital plexus using a capillary tube coated with heparin. For the sparse sampling, blood samples at 0.25, 1, 4, and 24 h were collected from 4 mice and blood samples at 0.5, 2, and 8 h were collected from the other 4 mice. The blood samples were centrifuged at 13,000 rpm for 10 min to obtain a supernatant plasma, which was stored at –80°C until LC-MS/MS analysis.

One hundred fifty microliters of berberine methanol solution (1 ng/mL, IS) were added to 30 μL of plasma samples and QC samples, mixed vigorously for 5 min, and centrifuged at 16,000 × g for 5 min. Supernatants (100 μL each) were transferred to the autosampler vials and 5 μL aliquots were injected into the LC-MS/MS system.

The anti-cancer activity of CUR on B16F10 bearing mice

Male C57BL/6 mice (7 weeks, 23-25 g, totally 24 mice) were randomly divided into four groups. B16F10 cells were subcutaneously inoculated into the right flank of the mice (3 × 10⁶ cells in 150 µL PBS per mouse). After 1 week of B16F10 cell inoculation, the mice were injected intraperitoneally with CUR, CUR-SMB7, or CUR-PP85-SMB7 (20 mg CUR equivalent /kg for 7 days). The tumor volume and body weight of the B16F10-bearing mice were measured and recorded once a day for 1 week. Tumor volume (V) was calculated from the length (L) and width (W) of the tumor (V = L × W²) at the indicated time points. On seventh day, B16F10-bearing mice received the final CUR dose, and the tumor volume and body weight of the mice were measured. At the end of the experiment (4 h after the final dose), the animals were sacrificed and the plasma, tumors, heart, lung, kidney, liver, and pancreas were harvested for weighing. Tissue samples were homogenized with 4 volumes of saline and 50 µL of tissue homogenates were stored at –80°C until LC-MS/MS analysis.

One hundred fifty microliters of berberine methanol solution (1 ng/mL, IS) was added to 30 μL of plasma samples or 50 μL of tissue homogenate samples, mixed vigorously for 5 min, and centrifuged at 16,000 × g for 5 min. Supernatants (100 μL each) were transferred to the autosampler vials and 5 μL aliquots were injected into the LC-MS/MS system.

LC-MS/MS analysis of CUR

CUR concentrations were analyzed using an Agilent 6430 triple quadrupole mass spectrometer equipped with an Agilent Infinity 1260 Infinite II high performance liquid chromatography (Agilent Technologies, Santa Clara, CA, USA) according to the previously established method by Song et al. with slight modification (Song et al. 2016). Separation was performed on a Sunniest C18 column (2.0 mm × 100 mm, 3 μm; Chromanik, Osaka, Japan) using a mobile phase consisting of water and methanol (20:80 v/v) containing 0.1% formic acid at a flow rate of 0.2 mL/min. Quantification was carried out using multiple reaction monitoring (MRM) at m/z 369.2→285.1 for CUR and m/z 360.0→320.0 for berberine in positive ionization mode. The lower limit of quantification (LLOQ) was determined to be 1 ng/mL and the standard curve exhibited linearity over the range of 1-10,000 ng/mL. Intra- and inter-day precision and accuracy had coefficients of variance of less than 15%.

Data analysis and statistics

The WinNonlin 5.1 software (Pharsight Co., Mountain View, CA, USA) was utilized to calculate the pharmacokinetic parameters. All data are expressed as the mean ± standard deviation (SD). Mann–Whitney U-test for non-parametric variables was used for statistical analysis via SPSS for Windows, version 25.0 (IBM Corp., Armonk, NY, USA).

Effect of SMB7 and PP85 on CUR solubility and P-gp-mediated efflux

CUR accumulation in LLC-PK1-P-gp cells was increase by the presence of CsA, a representative inhibitor of P-gp (Kwon et al. 2020). Conversely, the addition of SMB7 did not increase the cellular accumulation of CUR, suggesting that SMB7 did not modulate P-gp function. PP85 increased cellular CUR concentration in a concentration dependent manner and significantly higher than that of TPGS (Fig. 1A). Since we previously formulate solid dispersion of CUR: TPGS: mannitol with a ratio of 1:10:10 (w/w/w) and it also showed P-gp modulating effect and enhanced anti-tumor effect (Song et al. 2016), we compared the P-gp-modulating effect of PP85 with TPGS.

Figure 1. (A) Cellular accumulation of 5 µM CUR in the presence of CsA (20 µM), TPGS (1-1,000 µg/mL), PP85 (1-1,000 µg/mL), or SMB7 (10-10,000 µg/mL) in LLC-PK1-P-gp cells. (B) The solubility of CUR is determined by increasing the ratio of PP85. (C) The solubility of CUR is determined by increasing the ratio of SMB7 in the presence of CUR:PP85 = 1:2 (w/w). Each bar represents the mean ± SD (n = 3).

CUR solubility was increased with increasing ratio of PP85 and the greatest increase was shown the ratio of PP85 to CUR was between 1:1 and 1:2 (w/w) (Fig. 1B). Based on the CUR solubility, CUR loading content in the formulation, and P-gp-mediated efflux, the preferred composition of CUR and PP85 was determined to be at a ratio of 1:2 (w/w). The CUR solubility in CUR-and PP85-loaded mesoporous SMB7 was gradually increased and the highest solubility was obtained when the ratio of CUR:PP85:SMB7 was 1:2:4 (w/w/w), which could be the final composition of CUR-and PP85-loaded mesoporous SMB7 formulation.

Preparation of CUR- and PP85-loaded silica nanoparticles (CUR-PP85-SMB7)

Since CUR showed low aqueous solubility (4.17 µg/mL) but non-ionic surfactant pluronic P85 can be dispersed in water (Song et al. 2016; Kwon et al. 2020), we applied a two- step method to prepare the CUR- and PP85-loaded silica nanoparticles. First, CUR was dissolved in a methanol:acetone = 1:1 (v/v) mixture and incorporated into mesoporous nanosilicates (CUR-SMB7) by the solvent evaporation method. Second, pluronic P85 dispersed in water was loaded into CUR-loaded mesoporous nanosilicates (CUR-PP85-SMB7) using a freeze-drying method. The prepared CUR-SMB7 and CUR-PP85-SMB7 were compared with CUR and SMB7. Representative photographs of each preparation were shown in Fig. 2A. Both CUR-SMB7 and CUR-PP85-SMB7 had a powdery appearance that was similar to SMB7, and in an orange color. CUR loading of CUR-SMB7 and CUR-PP85-SMB7 were determined as 15.0 ± 2.14% and 11.9 ± 0.37%, respectively. EE of CUR in CUR-SMB7 and CUR-PP85-SMB7 were determined as 75.1 ± 10.7% and 83.2 ± 2.62%, respectively.

Figure 2. (A) Representative photos and (B) Scanning electron microscope (SEM) images of CUR, SMB7, CUR-SMB7, and CUR-Pluronic P85-SMB7 (CUR-PP85-SMB7). The number indicates the magnification of each image and the scale bar represents 1 µm. (C and D) Bio-TEM images of CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 in dispersed in simulated body fluid (SBF, pH 7.4) at 0 h (C) and after the incubation for 48 h (D). The number indicates the magnification of each image and the scale bar represents 200 mm.

Fig. 2B shows the surface morphology of CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7 monitored by SEM. CUR has an irregularly sized square shape and SMB7 has spherical shapes. CUR-SMB7 and CUR-PP85-SMB7 both exhibited the same spherical shape as SMB7. The surface shape of CUR (irregular and rectangular) could not be observed in CUR-SMB7 and CUR-PP85-SMB7. These results indicate that CUR may have infiltrated into the SMB7. Subsequently, we monitored the morphological change of SMB7, CUR-SMB7, and CUR-PP85-SMB7 in SBF (pH 7.4) using a Bio-TEM (Baino and Yamaguchi 2020). As shown in Fig. 2C, SMB7, CUR-SMB7, and CUR-PP85-SMB7 are all spherical in shape, with small pores in the SBF buffer, initially. However, the spherical shapes of SMB7, CUR-SMB7, and CUR-PP85-SMB7 degraded after 48 h of incubation (Fig. 2C vs. 2D).

We next investigated the physicochemical properties of the formulation using XRD, DSC, and FT-IR analysis (Fig. 3). CUR exhibited sharp peaks in a 2 theta range of 5-30, indicating a typical crystalline structure (Fig. 3A). This result was consistent with the finding reported previously (Song et al. 2016). The diffraction pattern of SMB7, CUR-SMB7, and CUR-PP85-SMB7 show wide peaks at about 25, and similar to each other (Fig. 3A). The DSC curve of CUR and SMB7 featured a sharp and wide endothermic peak at around 173°C and 88°C, respectively, indicating their crystalline nature. The DSC pattern of CUR-SMB7 and CUR-PP85-SMB7 showed similar but smaller patterns compared with SMB7, with disappeared CUR peaks (Fig. 3B). Next, each formulation was analyzed using FT-IR. CUR exhibited many peaks at wavelengths from 400 to 1,500 and 3,500 cm^-1, while SMB7 exhibited peaks at 400 and 1,000 cm^-1. PP85 also showed multiple peaks around the 1,200-1,500 cm^-1. However, CUR-SMB7 and CUR-PP85-SMB7 had the same wavelength peaks as SMB7, even though CUR and PP85 were loaded (Fig. 3C). Taken together, the XRD, DSC, and FT-IR patterns of CUR-SMB7 and CUR-PP85-SMB7 suggested that CUR and PP85 were fully embedded into SMB7 and, thus, showed similar patterns compare with those of SMB7.

Figure 3. (A) The X-ray diffraction (XRD) patterns, (B) differential scanning calorimetry (DSC) thermograms, and (C) Fourier-transform infrared spectroscopy (FT-IR) spectrometry of CUR, SMB7, CUR-SMB7, and CUR-PP85-SMB7. (D) CUR solubility of CUR alone, CUR-SMB7, and CUR-PP85-SMB7 in distilled water and simulated body fluid (SBF, pH 7.4). (E) In vitro drug release profile of CUR from the cellulose bag containing CUR, CUR-SMB7, and CUR-PP85-SMB7 for 72 h. *p < 0.05, **p < 0.01, and ***p < 0.001 compared with CUR group; ⁺p < 0.05, ⁺⁺p <0.01, ⁺⁺⁺p < 0.001 compared with CUR-SMB7 group. Data are expressed as mean ± SD (n = 3).

The aqueous solubility of CUR, CUR-SMB7, and CUR-PP85-SMB7 were measured in distilled water and SBF (pH 7.4). As shown in Fig. 3D, the CUR solubility in CUR-SMB7 and CUR-PP85-SMB7 was 95~154-fold and 272~257-fold increased at distilled water and simulated body fluid, respectively. The results suggested that incorporation of PP85 with CUR into SMB7 significantly increased the CUR solubility compared with CUR-SMB7.

CUR showed a very poor drug release at 2% over a 72-h period (Fig. 3E). However, the formulations with PP85 and SMB7 both showed a different pattern. CUR from CUR-SMB7 and CUR-PP85-SMB7 began to release after 4 h and showed a steady increase over 48-72 h. Additionally, the amount of drug release differed depending on the presence or absence of PP85. As a result, the CUR released from CUR-SMB7 and CUR-PP85-SMB7 were 53% and 85%, respectively, at 72 h. These results suggest that the CUR release pattern were greatly improved due when incorporated with SMB7 and PP85.

Cytotoxicity of CUR, CUR-SMB7, CUR-PP85-SMB7 on B16F10 cell

The cytotoxicity of CUR, SMB7, CUR-SMB7 and CUR-PP85-SMB7 was investigated in B16F10 cells in concentrations ranging from 0.1-100 µM CUR or the corresponding dose. The half maximal inhibitory concentration (IC₅₀) of CUR was found to be 1.1 µM. Compared to the IC₅₀ of CUR, the IC₅₀ of CUR-SMB7 and CUR-PP85-SMB7 decreased to 0.52 µM and 0.38 µM, respectively (Fig. 4A). The cytotoxicity of SMB7 to B16F10 cells was 1,343 µg/mL (Fig. 4B). The high IC₅₀ values of SMB7 suggested that significant cytotoxicity may not be caused by the mesoporous nanosilicates as the vehicle.

Figure 4. (A) The effect of CUR (●), CUR-SMB7 (○), and CUR-PP85-SMB7 (▼) on B16F10 cell viability, expressed as molecular concentrations of curcumin in B16F10 melanoma cells. (B) The effect of SMB7 on B16F10 cell viability. Data represent the means ± SD (n = 3).

Pharmacokinetics of CUR in mice following the administration of CUR, CUR-SMB7, and CUR-PP85-SMB7

Fig. 5 and Table 1 show the pharmacokinetic properties and parameters of CUR, CUR-SMB7, and CUR-PP85-SMB7 after intraperitoneal administration of 5 mg/kg of CUR. Consequently, all three formulations reached their highest peak immediately after administration and gradually decreased over 24 h. Pharmacokinetically, Significant differences among pharmacokinetic parameters of CUR, CUR-SMB7, and CUR-PP85-SMB7 were observed in C_max, AUC_24h, and AUC∞ (Table 1).

Figure 5. Plasma concentration-time profile of CUR following the intraperitoneal administration of CUR (●), CUR-SMB7 (○), and CUR-PP85-SMB7 (▼) at a curcumin equivalent dose of 5 mg/kg in C57BL/6 mice. Data represent the means ± SD (n = 4).

Table 1 . The pharmacokinetic parameters of CUR following the intraperitoneal administration of CUR, CUR-SMB7, and CUR-PP85-SMB7 in C57BL/6 mice.

Group	CUR	CUR-SMB7	CUR-PP85-SMB7
Cmax (ng/mL)	72.5 ± 11.1	328 ± 108.9*	345.4 ± 165.2*
Tmax (h)	0.3 ± 0.1	0.3 ± 0.1	0.3 ± 0.0
AUC24h (ngxh/mL)	270.8 ± 130.5	668.6 ± 287.1	1,120.5 ± 525.2*
AUC∞ (ngxh/mL)	677.8 ± 351.1	1,006.3 ± 245.5	1,740.6 ± 321.4*,+
T1/2 (h)	15.8 ± 7.3	19.3 ± 11.6	13.4 ± 7.3
MRT (h)	8.3 ± 3.4	7.9 ± 1.9	9.9 ± 2.5

C_max, maximum plasma concentration; T_max, time to reach C_max; AUC_24h, area under plasma concentration-time curve from zero to 24 h; T_1/2, elimination half-life; MRT, mean residence time..

Data were expressed as mean ± SD (n = 4)..

*p < 0.05 compared with CUR group; ⁺p < 0.05 compared with CUR-SMB7 group..

C_max of CUR was found to be 72.5 ± 11.1 ng/mL. This level was significantly increased in the CUR-SMB7 and CUR-PP85-SMB7 groups. C_max of CUR-SMB7 and CUR-PP85-SMB7 increased 4.5-fold and 4.75-fold compared with that of the CUR group (Table 1). In addition, AUC_24h and AUC∞ of CUR-SMB7 were significantly higher than those of CUR and the fabrication of PP85 also significantly contribute to the higher AUC_24h than CUR-SMB7 (Table 1). Contrary to the exposure-related parameters, time-related parameters, such as time to reach maximum concentration (T_max), elimination half-life (T_1/2), and mean residence time (MRT) of CUR were not changed by the incorporation into nanosilicates or the addition of PP85.

Anti-cancer activity of CUR on B16F10 bearing mice

To investigate the anti-cancer activity of CUR, the mice were injected intraperitoneally with CUR, CUR-SMB7, or CUR-PP85-SMB7 (all doses were equivalent to 20 mg/kg) for 7 days, and the tumor volume and body weight of the mice were monitored. The tumor volume of the control mice (i.e., B16F10-bearing mice in the vehicle treatment group) increased with time. Notably, CUR treatment significantly reduced the tumor volume compared with the control. The incorporation of CUR into nanosilicates and the addition of PP85 (CUR-SMB7, CUR-PP85-SMB7) appeared to significantly inhibit tumor growth compared to the CUR group (Figs. 6A and 6B). That is, tumor volume and tumor weight following CUR, CUR-SMB7, and CUR-PP85-SMB7 significantly decreased compared with the control group. The tumor volume and tumor weight from the CUR-SMB7 and CUR-PP85-SMB7 groups significantly reduced compared with the CUR group (Figs. 6A and 6B). Moreover, the CUR concentration in the tumor tissues was the highest in the CUR-PP85-SMB7 group, followed by the CUR-SMB7 and the CUR groups, which seemed to be consistent with the tumor growth inhibition (Fig. 6C). However, there was no clear differences in other organs (heart, lung, spleen, kidney, and liver) or the body weight of the mice (Figs. 6D and 6E). The results suggest that CUR, CUR-SMB7, and CUR-PP85-SMB7 treatments are unlikely to result in general toxicity during a 7-day treatment period.

Figure 6. (A) Tumor growth trend of B16F10 melanoma bearing C57BL/6 mice following repeated intraperitoneal administration of CUR, CUR-SMB7, and CUR-PP85-SMB7 at a CUR equivalent dose of 20 mg/kg for 7 days and compared with vehicle treated control group. (B) Tumor tissue weight, (C) CUR concentration in tumor tissue, (D) total body weight, and (E) organ weight of B16F10 melanoma bearing C57BL/6 mice after the last dose of vehicle, CUR, CUR-SMB7, and CUR-PP85-SMB7 on the seventh day. *p < 0.05, **p < 0.01, and ***p < 0.001 compared with CUR group; ⁺p < 0.05, ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001 compared with CUR-SMB7 group. Data represent the means ± SD (n = 6).

We also measured the plasma and tissue concentrations of CUR at 4 h after the last dose of CUR, CUR-SMB7, and CUR-PP85-SMB7 (Fig. 7). Free CUR was highly distributed to the spleen and kidney, followed by the lung and liver. The CUR concentration in the heart was lower than in the plasma after the repeated intraperitoneal administration of CUR. However, plasma CUR concentration and its distribution into the spleen and liver was increase by CUR-SMB7 and CUR-PP85-SMB7 compared with free CUR. However, tissue distribution of CUR between CUR-SMB7 and CUR-PP85-SMB7 was not different (Fig. 7).

Figure 7. Tissue concentrations of CUR at 4 h after the last dose of repeated intraperitoneal administrations of CUR, CUR-SMB7, and CUR-PP85-SMB7 in C57BL/6 mice bearing B16F10 melanoma cells. *p < 0.05 compared with CUR group; +p < 0.05 compared with CUR-SMB7 group. Data are expressed as mean ± SD (n = 6).

CUR, the active ingredient extracted from turmeric, has recently been reported as a promising treatment for melanoma (Zhang et al. 2015; Mirzaei et al. 2016). Previously, we increased the bioavailability of CUR and sensitized the anti-cancer activity of paclitaxel by using solid dispersion formulation of CUR in combination with TPGS as a P-gp inhibitor and solubilizing agent (Song et al. 2016). However, the CUR solid dispersion was formulated with a ratio of 1:10:10 (w/w/w) of CUR:TPGS:mannitol, which demonstrated a low drug content of about 5%. To increase the drug content with increased CUR solubility and P-gp modulating efficacy, we tried to prepare a mesoporous nanosilicate formulation that included CUR and a P-gp inhibitor. SMB7, a newly developed mesoporous silicate nanoparticle from CEN CO., Ltd (Miryang-si, Gyeongnam, Republic of Korea), has a large surface area of 1,000 m²/g and an abundant pore volume of 1 cm³/g. It also can be used as a carrier for drug delivery due to its biocompatibility, biodegradability, and physicochemical stability (Xu et al. 2023). PP85 can increase CUR solubility and accumulate CUR in tumor cells by inhibiting P-gp-mediated efflux (Kwon et al. 2020). The inhibitory effect of PP85 on the P-gp-mediated efflux of CUR was greater than TPGS (Fig. 1A). Additionally, it can enhance anti-cancer effects and reduce the possibility of adverse effects, especially on the immune system (Batrakova et al. 1996; Melik-Nubarov et al. 1999; Yamagata et al. 2007).

Therefore, in this study, we prepared CUR- and PP85-loaded silica nanoparticles (CUR-PP85-SMB7) with a ratio of 1:2:4 (w/w) of CUR:PP85:SMB7 using a two-step method (i.e., solvent evaporation method, followed by the freeze drying method). CUR and PP85 successfully incorporated into the pore of SMB7, which was evidenced by the SEM images as well as the XRD, DSC, and FT-IR results (Figs. 2B and 3). The crystalline nature of CUR was changed into an amorphous state based on the XRD and DSC results (Figs. 3A and 3B). This may explain the increased solubility and the dissolution profile of CUR-PP85-SMB7 and CUR-SMB7 compared with CUR itself (Figs. 3D and 3E).

Next, we investigated the in vitro anti-cancer effect of CUR, CUR-SMB7, and CUR-PP85-SMB7 in B16F10 melanoma cells and obtained IC₅₀ values of 1.1 µM, 0.52 µM, and 0.38 µM, respectively (Fig. 4A). The increased solubility and release profile of CUR by using silicate nanoparticles and PP85 may contribute to the increased cytotoxicity of CUR in B16F10 melanoma cells. Additionally, the increased CUR concentration in B16F10 melanoma cells from the P-gp modulation by PP85 can also contribute to the increased cytotoxicity. Considering the loading amount of CUR is about 11.9-15%, CUR-SMB7 and CUR-PP85-SMB7 at their IC₅₀ values contained 189 ng/mL of CUR and 1,072 ng/mL of SMB7, and 138 ng/mL of CUR and 1,014 ng/mL of SMB7, respectively. The IC₅₀ value of SMB7 in B16F10 cells was about 1000-fold higher than the amount of SMB7 used in the measurement of cell cytotoxicity. The results suggest that the cytotoxicity effect came from CUR and not from SMB7.

CUR-SMB7 and CUR-PP85-SMB7 also increased the C_max and AUC of CUR compared to CUR alone in mice following the intraperitoneal administration of CUR, CUR-SMB7, and CUR-PP85-SMB7, with all containing the same dose of CUR. These results also indicated that the plasma concentration of CUR was increased due to an increased solubility and increased release profile, as well as the protection from the degrading environment in the body by being embedded in SMB7 in addition to PP85.

Finally, the anti-cancer effect of CUR, CUR-SMB7, and CUR-PP85-SMB7 was evaluated in B16F10-bearing mice. B16F10 melanoma cells are known to express P-gp. The downregulation of P-gp and inhibition of P-gp activity could reduce tumor drug resistance (Yang et al. 2018; Ni et al. 2022). The CUR, CUR-SMB7, CUR-PP85-SMB7 groups were all found to inhibit tumor growth compared to the control group and the anti-cancer effect of CUR-SMB7 and CUR-PP85-SMB7 were better than that of CUR alone, which was consistent with the in vitro cytotoxicity and pharmacokinetic results. During the repeated treatment using CUR, CUR-SMB7, or CUR-PP85-SMB7 for 1 week, the organ weights and body weight did not change. Only the weight of the cancer tissues was reduced by CUR and by CUR-SMB7 or CUR-PP85-SMB7 treatment to a greater extent. These results may be attributed to the mechanism that inhibits tumor growth through a number of previously reported cell signaling pathways in melanoma cells, including CUR’s cell proliferation pathway, cell survival pathway, and apoptosis pathway (Zhang et al. 2015; Mirzaei et al. 2016). Being formulated as CUR-SMB7 and CUR-PP85-SMB7, CUR demonstrated improved anti-cancer effect. This can be explained by the enhanced pharmacokinetics properties of CUR as well as increased CUR accumulation in tumor tissues due to the P-gp modulation by PP85 (Figs. 5 and 6). This higher intratumoral CUR concentration was likely caused by the P-gp inhibition by PP85. The results suggest that CUR-SMB7 provided greater anti-cancer effect than CUR alone and, importantly, CUR-PP85-SMB7 demonstrated the best treatment outcome.

According to the previous reports, CUR was not highly distributed to the organs in rats and mice and, therefore, the CUR concentrations in the tissue was lower than plasma CUR concentration or below detection limit after intravenous injection of CUR (Ravindranath and Chandrasekhara 1980; Tsai et al. 2011; Wang et al. 2018). When administered as a CUR nanoformulation via tail vein, CUR distribution was increased in the spleen, liver, lung, and kidney (Tsai et al. 2011). In this study, CUR was readily distributed into spleen and kidney whereas CUR concentration in heart, lung, and liver were similar or lower than plasma concentration via the repeated intraperitoneal administration. By using CUR-SMB7 and CUR-PP85-SMB7 formulation, CUR distribution to the spleen, lung, and liver was increased (Fig. 7), which is consistent with the previous report (Tsai et al. 2011). Therefore, the CUR concentration in the spleen, liver, lung, and kidney were much higher than that in tumor tissue (Fig. 7). Although the tumor tissue was significantly reduced by the CUR-SMB7 and CUR-PP85-SMB7 treatment, other tissues did not show significant morphological change. It suggested that the specific anti-cancer mechanism existed in B16F10 melanoma cells. In addition, the CUR concentration in the cancer tissue at 4 h of CUR dosing were 0.20-0.58 µM, which was similar to IC₅₀ values in B16F10 melanoma cells calculated from in vitro cell viability test (Fig. 4A).

In conclusion, CUR-PP85-SMB7 showed better solubility and superior drug release than CUR alone, leading to increased plasma exposure and tumor accumulation. Consequently, superior anti-cancer activity against B16F10 cells was observed relative to CUR alone. The use of PP85 as solubility enhancer and P-gp modulator and the use of SMB7 as nano-carrier could potentially enhance the anti-cancer efficacy of CUR in B16F10 melanoma cells.

The authors declare that they have no conflict of interest.

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2020R1I1A3074384 and NRF-2020R1A5A2017323).