γ-Cyclodextrin metal–organic framework as a carrier to deliver triptolide for the treatment of hepatocellular carcinoma

Triptolide (TPL) has been employed to treat hepatocellular carcinoma (HCC). However, the poor water solubility of TPL restricts its applications. Therefore, we prepared TPL-loaded cyclodextrin-based metal–organic framework (TPL@CD-MOF) to improve the solubility and bioavailability of TPL, thus enhancing the anti-tumor effect on HCC. The BET surface and the pore size of TPL@CD-MOF were 10.4 m2·g−1 and 1.1 nm, respectively. The results of XRD indicated that TPL in TPL@CD-MOF was encapsuled. TPL@CD-MOF showed a slower release than free TPL in vitro. Moreover, the CD-MOF improved the bioavailability of TPL. TPL@CD-MOF showed slightly higher, but statistically significant, anti-tumor efficacy in vitro and in vivo compared to free TPL. In addition, TPL@CD-MOF exhibited a modest improvement of the anti-tumor effects, which may be associated to the enhanced in vivo absorption. Overall, these findings suggested the potential CD-MOF as oral drug delivery carriers for anti-tumor drugs. The process of TPL loading into CD-MOF and its enhanced oral bioavailability and anti-tumor activity.


Introduction
Hepatocellular carcinoma (HCC) accounts for the majority of primary liver cancers. Liver cancers are the fourth most common cause of cancer-related mortality worldwide [1]. Surgical resection is still the preferred method for the curative treatment of HCC. However, HCC is usually diagnosed at a late stage with only approximately 15% of HCC patients eligible for operative treatment [2][3][4][5]. In addition, the effects of first-line therapies with sorafenib or lenvatinib and second-line therapies with regorafenib, cabozantinib, or ramucirumab are not so satisfactory, which reminds us to develop new chemotherapy drugs to improve the efficacy of HCC.
Triptolide (TPL), extracted from Tripterygium wilfordii, has a variety of bioactivities, such as antioxidation, antiinflammation, and anti-cancer effects [6]. For anti-tumor, TPL has been employed in the treatment of leukemia lung cancer, colon cancer, and HCC [5,[7][8][9][10]. However, the clinical applications of TPL are limited due to its narrow therapeutic window, severe toxicity, and poor water solubility [7,11]. Drug delivery systems (DDSs) have been widely used to improve solubility of insoluble drugs and reduce drug toxicity [12]. In recent decades, various DDSs have been developed to deliver TPL, such as solid lipid nanoparticles, liposomes, polymer micelles, and microemulsion. However, the drug loadings of these DDSs were not so satisfactory [13]. Therefore, it is necessary to develop a new DDS to increase the drug solubility, improve the drug loading and enhance the therapeutic effect. Among different DDSs, γ-cyclodextrin-based metal-organic framework (γ-CD-MOF) with the characteristics of adjustable structure, high specific surface area, good biocompatibility, and high drug loading has received much attention in recent years and has been successfully applied in drug delivery [14]. Drugs with poor water solubility such as oleanolic acid, curcumin, ibuprofen, and honokiol have been encapsulated into CD-MOF. The solubilities of the drugs were enhanced and the bioavailabilities were also improved [12,[15][16][17].
Inspired by the researches mentioned above, we have prepared TPL-loaded CD-MOF (TPL@CD-MOF) to improve the solubility and bioavailability of TPL, thus enhancing the anti-tumor effect on HCC. We hypothesized that after encapsulating TPL into CD-MOF, the solubility and bioavailability of the drug could be improved. TPL@CD-MOF was characterized by scanning electron microscopy (SEM), nitrogen sorption-desorption experiment and X-ray diffraction (XRD). Drug loading, in vitro release profile, celluptake study, and in vitro cytotoxicity of TPL@CD-MOF were tested. Moreover, in vivo pharmacokinetics and pharmacodynamics were also investigated.
The Huh-7 cell line and Caco-2 cell line were purchased from the Cell Bank of Typical Culture Preservation Committee of the Chinese Academy of Sciences (Shanghai, China). Huh-7 cells and Caco-2 cells were cultured in DMEM supplemented with 10% FBS, 1% penicillin/streptomycin in a humidified incubator with 5% CO 2 at 37 °C.
Healthy male Sprague-Dawley (SD) mice (200 ± 20 g) and Balb/c-nu mice (18 ± 2 g) were randomly assigned to different groups. The experiment was approved by the Ethics Committee of Ninth People's Hospital, affiliated with Shanghai Jiao Tong University School of Medicine before the research.

High-performance liquid chromatography (HPLC) analysis
The detection of TPL was performed on a Waters e2695 HPLC system (Waters Technologies, USA) with an Agilent TC-C18 column (250 mm × 4.6 mm, 5 μm) and acetonitrile/water = 70:30 as the eluent. The flow rate was set at 1.0 mL·min −1 , and the column temperature was set at 30 °C. The detection wavelength was 218 nm with an injection volume of 20 μL.

Synthesis and activation of CD-MOF
The CD-MOF was synthesized according to the previously published protocols [17]. Briefly, γ-CD (3.24 g) and potassium hydroxide (1.12 g) were dissolved in 100 mL distilled water with a 20min incubation at 50 ℃. Then, the obtained solution was mixed with 60 mL methanol. Thereafter, 160 mL PEG 20,000 methanolic solution (8 mg·mL −1 ) was added and the solution was kept in cold water for about 12 h to collect the precipitates. To remove the residual potassium hydroxide, γ-CD, and PEG 20,000, the precipitates were washed 3 times with ethanol (60 mL). To activate the obtained CD-MOF, the precipitates were immersed in dichloromethane for 72 h (refreshing dichloromethane every 24 h). At last, the blank CD-MOF was obtained by centrifugation and dried under vacuum at 50 ℃ overnight.

Drug loading
TPL was loaded into CD-MOF with solvent adsorption method [18]: 80 mg TPL was dissolved in 2 mL acetonitrile under ultrasonication. 20 mg Blank CD-MOF were then immersed in the TPL solution, stirring for 24 h on a magnetic stirrer (500 rpm) at room temperature. The TPL@CD-MOF was collected by centrifugation, followed by washing 3 times with acetonitrile and drying for 24 h under vacuum at room temperature. C6@CD-MOF was prepared for the cellular uptake study. For C6 loading, 2 mg C6 was dissolved in 2 mL acetonitrile under ultrasonication. Other steps were the same as mentioned above.
To determine the weight of TPL loaded in the collected TPL@CD-MOF, the samples were dissolved in acetonitrile and TPL was extracted from the precisely weighted TPL@CD-MOF with sonication and then centrifuged (10,000 rpm, 10 min). The process was repeated three times to ensure the complete extraction. TPL in acetonitrile was quantified using HPLC. The loading capacity of TPL was calculated with the following equation [14]: whereW TPL and W TPL@CD-MOF are weights of TPL and TPL@ CD-MOF, respectively.

Characterization
The surface morphology of the blank CD-MOF and TPL@ CD-MOF was characterized with a Hitachi S-4800 Scanning Electron Microscopy (SEM, Tokyo, Japan). The

DL = W TPL W TPL@CD−MOF
nitrogen adsorption-desorption isotherms of CD-MOF and TPL@CD-MOF at 77 K were determined by a Quadrasorb 2 analyzer. X-ray diffraction (XRD) patterns of the free TPL, physical mixture (PM), blank CD-MOF, and TPL@ CD-MOF were conducted on a D8 Advance X-ray diffractometer (Karlsruhe, Germany).

Equilibrium solubility and in vitro release
Solubilities of the pure TPL and TPL@CD-MOF were determined in pure water with the published method [14]. In brief, an excess amount of pure TPL or TPL@CD-MOF was added into pure water, shaking (500 rpm) at 25 °C for 72 h to reach the equilibrium. After centrifugation (10,000 rpm), the supernatant was analyzed by HPLC to measure the concentration of TPL in the solutions.
The in vitro release behavior of the free TPL and TPL@ CD-MOF was evaluated. In brief, 1 mg TPL and equivalent TPL@CD-MOF were dispersed in 1 mL of release media and placed in cellulose membrane dialysis bag (molecular cutoff, 8-14 kDa) with 100 mL release media (1% sodium dodecyl sulfate solution, pH 1.2, 7.4, and 6.8). To simulate the fate of TPL@CD-MOF after oral administration, release media with pH 1.2, 7.4, and 6.8 were employed to simulate the gastric fluid, the intestinal fluid, and the colonic fluid, respectively. The paddle speed was set at 100 rpm, and the temperature was set at 37 ℃. 1 mL release media was then withdrawn at predetermined time points and analyzed by HPLC. In the meantime, 1 mL fresh medium was replenished.

Cell uptake studies
The cell uptake of C6@CD-MOF by Caco-2 cells was evaluated by a Leica TCS-SP8 confocal laser scanning microscopy (CLSM, Wetzlar, Germany). 1 mL Caco-2 cells was seeded into confocal dishes at the density of 5 × 10 4 ·mL −1 12 h prior to the experiment. Then, the culture medium was removed and the cells were washed twice with PBS and incubated with free C6 or C6@CD-MOF solution for 2 h. Thereafter, the cells were rinsed three times with PBS and fixed with 4% paraformaldehyde. After being fixed, the cells were washed three times with PBS and stained with Hoechst 33258 for 5 min. Finally, the cells were washed three times with PBS again and observed by CLSM.

Cell viability assay
CCK-8 was utilized to investigate the cell viability after the treatment with free TPL and TPL@CD-MOF. In brief, 150 μL Huh-7 cells were seeded in 96-well plates at the density of 5 × 10 4 ·mL −1 overnight. Then, the cells were treated with TPL or TPL@CD-MOF. After treatment for 48 h, 20 μL CCK-8 was added into each well incubating for 2 h. At last, an Infinite M200 PRO microplate reader (Männedorf, Switzerland) was used to assess cell viability by measuring the absorbance at 450 nm. The 50% inhibitive concentration (IC 50 ) was performed with CompuSyn software (Biosoft, UK).

In vivo pharmacokinetic studies
Ten healthy male SD mice (200 ± 20 g) were randomly divided into two groups with five mice in each group. They were orally administered with free TPL or TPL@CD-MOF at a dose of 1.5 mg·kg −1 . Blood samples were collected from the retro-orbital plexus into heparinized tubes at predetermined time points centrifuging immediately (3000 rpm for 10 min) to obtain plasma. Then, 0.1 mL plasma was extracted with 1.2 mL ethyl acetate for 3 times. The top organic layers were combined and dried. The residue was redissolved in 0.1 mL of acetonitrile and centrifuged. The supernatant was analyzed by HPLC.

In vivo anti-tumor activity
For establishing xenografts, 2 × 10 6 Huh-7 cells were injected subcutaneously into the left axillary region of male Balb/c-nu mice [5]. The tumor volumes and weights of the mice were measured every 2 days. When the tumor volume grew up to 50 mm 3 [19], the mice were then randomized to 3 groups with 5 mice in each group and administered by gavage with saline, free TPL or TPL@CD-MOF at a dose of 1.5 mg·kg −1 TPL everyday for 10 days. All mice were sacrificed one day after the last administration. The tumors were harvested and weighed, followed by TUNEL and hematoxylin and eosin (H&E) staining assay.

Statistical analysis
Data were expressed as mean ± standard deviation (SD). The pharmacokinetic parameters were analyzed with DAS 2.0 software (BioGuider Co, Shanghai, China). All data analysis of variance were performed with the SPSS version 19.0 software (SPSS Inc., Chicago, USA). P ≤ 0.05 was considered statistically significant.

Preparation and characterization of TPL@CD-MOF
CD-MOF was synthesized with solvothermal method, which was identified as one of the best procedures for the synthesis of CD-MOF. The drug loading of TPL@ CD-MOF (28.89%) was significantly higher than that of other TPL loaded nanoparticles [20]. As shown in Fig. 1a, SEM images of the blank CD-MOF and TPL@CD-MOF indicated that the particles were uniform cubic crystals with a particle size of about 200-500 nm. The nitrogen adsorption-desorption isotherms of the TPL@CD-MOF and blank CD-MOF were shown in Fig. 1b. The BET surface and the pore size of blank CD-MOF and TPL@CD-MOF were 1134.5 m 2 ·g −1 and 10.4 m 2 ·g −1 , and 1.6 nm and 1.1 nm, respectively. The decreases of the BET surface and the pore size reflected the success of drug loading, which was similar to other drug-loaded CD-MOF [17]. XRD was used to confirm the physical state of TPL in TPL@CD-MOF. As shown in Fig. 1c, the characteristic peaks of XRD patterns were almost the same for CD-MOF in the absence and presence of TPL before and after drug loading. The free TPL and PM had characteristic diffraction peaks at 2θ = 8.60°, 15.26°, and 33.80°. After loading into TPL@CD-MOF, the characteristic peaks of TPL disappeared, which indicated that TPL in TPL@CD-MOF was encapsuled and converted from crystalline state into amorphous state.

Equilibrium solubility and in vitro release
The solubility of drug is a key parameter and determines its bioavailability. The clinical application of TPL has been limited due to its poor solubility [17]. After loading into CD-MOF, the equilibrium solubility of TPL loaded into TPL@CD-MOF in pure water reached 308.19 μg·mL −1 , which was about 9.5 times higher than that of free TPL (Fig. 2a). The enhanced solubility of TPL might be attributed to TPL in TPL@CD-MOF transformed from crystalline state into amorphous state, as shown in XRD. In vitro drug release of TPL and TPL@ CD-MOF was carried out in release media with pH 1.2, 7.4, and 6.8 (containing 1% sodium dodecyl sulfate solution). In the release media with pH 1.2, the free TPL and TPL@CD-MOF showed a burst release with about 40.00% of the drug released in the first 2 h (Fig. 2b). After an additional 4 h release in the simulated intestinal fluid (pH 7.4), the free TPL and TPL@CD-MOF released 82% and 72%, respectively. For the release in pH 6.8, the cumulative release percentage of both free TPL and TPL@CD-MOF reached about 97% within 12 h. On the whole, TPL@CD-MOF showed a slower release than free TPL, which may be attributed to the drug diffusion being hindered after loading into CD-MOF.

Cell uptake studies
For cell uptake studies, Caco-2 cells were adopted to simulate the gastrointestinal (GI) drug barrier for oral delivery [21]. C6 was used as a fluorescent probe to label CD-MOF. As shown in Fig. 3, C6@CD-MOF had a higher green fluorescent signal level in cytoplasm than free C6. The results were in accordance with the previous literature and accounted for a higher diffusion and absorption of C6 into Caco-2 cells [17]. Based on the results, it could  Fig. 4 The cytotoxicity of free TPL and TPL@CD-MOF on Huh-7 cells be inferred that CD-MOF potentially improved the cell internalization of drugs at the GI level, a previous and necessary step for drugs into the bloodstream.

Cell viability assay
The cytotoxicity of free TPL and TPL@CD-MOF on Huh-7 cells were evaluated by CCK-8 assay. As shown in Fig. 4, a dose-dependent increase in cell death was observed after 48 h of incubation with free TPL or TPL@CD-MOF. The IC 50 of free TPL and TPL@CD-MOF on Huh-7 cells were    14.02 and 7.80 nM, respectively. The TPL@CD-MOF were found to be more toxic towards Huh-7 cells compared with free TPL, which indicated that the TPL@CD-MOF helped drugs to make use of its anticancer efficacy

In vivo pharmacokinetic studies
The plasma concentration-time profiles of TPL after administering intragastrically with free TPL and TPL@CD-MOF in SD rats are shown in Fig. 5. The pharmacokinetic parameters are listed in Table 1. The bioavailability of TPL were significantly improved after loading into CD-MOF with C max and AUC 0-∞ of TPL@CD-MOF 1.62-and 1.82-fold improved respectively in comparison with those of the free TPL. Based on the improved bioavailability of TPL and cell-uptake, it could be included that CD-MOF improved the cell internalization of TPL at the GI level and made more TPL absorbed into the bloodstream. The T max for free TPL and TPL@CD-MOF was 0.5 h and 1.2 h, respectively. The longer T max of TPL@CD-MOF could be due to the slower release of TPL compared with that of the free TPL.

In vivo anti-tumor activity
Body weights (Fig. 6a) and tumors volumes (Fig. 6b) of the mice were measured every 2 days after the first dose. During the treatment, the weight of the mice in all groups grew slowly. TPL@ CD-MOF group showed superior anti-tumor efficacy with smaller tumor size and tumor weight (Fig. 6c, d) when compared with saline or free TPL group. Furthermore, we also evaluated the cell apoptosis in the harvested tumor tissues by H&E and TUNEL staining (Fig. 7). For TUNEL staining, TUNEL positivity (green fluorescence) was barely observed in the excised tumor tissues originated from saline group. However, TUNEL signals were observed in the excised tumor tissues originated from the free TPL group and TPL@CD-MOF group. Additionally, TUNEL signals were highest in TPL@CD-MOF group, which indicated that TPL@CD-MOF induced more apoptosis in tumor tissues when compared with other groups. Similar results were also observed in H&E staining assay. For saline group, tumor tissue displayed minor necrosis, compact structure and dense tumor cells. But various degrees of apoptotic morphological characteristics, like nuclear pyknosis and karyorrhexis were found in free TPL group and TPL@CD-MOF group, particularly obvious in TPL@CD-MOF group. From the above, TPL@ CD-MOF showed better anti-tumor effect in vivo compared to other groups, which might be attributed to the improved bioavailability of TPL@CD-MOF.

Conclusion
In this research, we prepared TPL@CD-MOF and evaluated the bioavailability and anti-tumor effect of TPL@CD-MOF in vitro and in vivo. CD-MOF with a particle size of about 200-500 nm were obtained. According to the results of nitrogen adsorption-desorption and XRD, we could conclude that TPL was successfully loaded into CD-MOF in an amorphous state. For equilibrium solubility and in vitro release, the solubility of TPL was improved in pure water and showed slower release behavior after loading into CD-MOF. Moreover, the CD-MOF improved the bioavailability of TPL. TPL@CD-MOF showed slightly higher, but statistically significant, anti-tumor efficacy in vitro and in vivo compared to free TPL, as shown by the changes in Huh-7 cell viability, tumor volumes, and H&E or TUNEL staining. Therefore, CD-MOF exhibited excellent potential as a carrier for insoluble anti-cancer drugs. Availability of data and materials All data are fully available without restriction.

Declarations
Ethics approval and consent to participate The animal experiment was approved by the Ethics Committee of Ninth People's Hospital, affiliated with Shanghai Jiao Tong University School of Medicine before the research (approval number SH9H-2020-A144-1).