A nanoparticulate pre-chemosensitizer for efficacious chemotherapy of multidrug resistant breast cancer

Small-molecule chemosensitizers can reverse cancer multidrug resistance (MDR), thus significantly improving the in vitro effect of chemotherapy drugs for MDR cancer cells, however, their in vivo effects are not always very good, because they are difficult to effectively accumulate in tumor and enter the same cancer with chemotherapy drugs after systemic administration due to individual biopharmaceutical properties. To overcome these limitations, here we study a novel nanoparticular pre-chemosensitizer which can be also used as nanocarrier of chemotherapy drugs. We take an ‘all in one’ approach to develop a self-assembled nanoparticle formula of amphiphilic poly(curcumin-dithiodipropionic acid)-b-poly(ethylene glycol)-biotin. The nanoparticle is capable of tumor-targeted delivery, responsive degradation at the intracellular level of glutathione and subsequent intracellular co-release of the chemosensitizer curcumin and the encapsulated chemotherapeutic drug doxorubicin to maximize a synergistic effect of chemosensitization and chemotherapy. We demonstrate that the antitumor efficacy of nanoparticle is much superior to that of doxorubicin in the multidrug resistant MCF-7/ADR xenografted nude mice.

including physical encapsulation of CUR in nanoparticles, chemical conjugation of CUR to hydrophilic compounds and the copolymerization of CUR with hydrophilic monomers, etc. For example, Shen group reported polycurcumin by condensation polymerization of curcumin with various hydrophilic PEGs 26 . Such polycondensates have the advantages of defined, high drug loading within the polymer backbones, leading to improved curcumin stability and tailored water-solubility. We designed and prepared Biotin-PEG-PCDA NPs through the use of a poly(active pharmaceutical ingredient) (PAPI) strategy where the API is incorporated into an intracellular cleavable polymer backbone in combination with the exploitation of self-assembly characteristics of amphiphilic di-block copolymers 28 Table S1 in the Supporting Information). The average molecular weights of PCDA-PEG-Biotin and PCDA-PEG were measured to be 13,820 and 16,010 Dalton, respectively. It is worth noting that the PCDA-PEG-Biotin has a high CUR loading capacity of 27.0 wt.% as well as > 500 fold greater . Subsequently, the DOX@PCDA-PEG-Biotin NPs are liable to escape from endosome/lysosome and then enter cytosol, where they will encounter relatively high level of GSH and thus decomposing into CUR followed by the release of encapsulated DOX (c). The released CUR can down-regulate the P-gp expression on MDR cancer cells and inhibit their ATP activity in favor of the intracellular and intranuclear accumulation of released DOX and CUR (d), playing a role as prechemosensitizer. water solubility than free CUR. We compared the stability of the PCDA-PEG-Biotin and CUR in phosphate buffered solution (PBS) at pH 7.4 by monitoring their UV absorption spectra ( Figure S5A and C). It can be found that free CUR quickly degraded, leading to an about 55% loss within 30 min ( Figure S5D), while no more than 5% of PCDA-PEG-Biotin decomposed after exposure for 24 h ( Figure S5B). This indicates that the PCDA-PEG-Biotin indeed stabilizes CUR from decomposition as expected. Further, the DOX@PCDA-PEG-Biotin NPs were prepared by a facile O/W emulsion-solvent evaporation method without using any other emulsifiers or surfactants. The obtained nanoparticles were characterized by transmission electron microscopy (TEM), dynamic laser scattering (DLS) and zeta potential measurements. The TEM results show that the DOX@PCDA-PEG-Biotin NPs are spherical ( Figure S6). The DOX@PCDA-PEG-Biotin NPs have a high DOX loading capacity of 11.9 wt.% with an encapsulation efficiency of 83.5%, possibly owing to the hydrophobic attraction, H-bonding and π -π stacking interaction between CUR and DOX ( Figure S1). The hydrodynamic diameter and zeta potential of the NPs are measured to be 82.1 nm and − 9.70 mV, respectively (Table S2).
GSH-triggered Degradation of Nano-carrier and Drug Release. The intracellular GSH concentration generally ranges from 0.5 to 10 mM, whereas the extracellular concentration of GSH is remarkably lower (only 2-20 μ M in plasma) 31 . We investigated the degradation of PCDA-PEG-Biotin NPs and the DOX release profiles of DOX@PCDA-PEG-Biotin NPs in PBS (pH 7.4) at two different concentrations of GSH (5 mM and 10 μ M). As shown in Fig. 3, the molecular weight of PCDA-PEG-Biotin almost did not change after 24 h incubation with 10 μ M GSH (Fig. 3A), but quickly reduced in the presence of 5 mM GSH (red and blue arrows in Fig. 3B). Correspondingly, the hydrodynamic diameter of the PCDA-PEG-Biotin nano-carrier remained almost unchanged at 10 μ M GSH within 3 h, but rapidly reduced after incubation with 5 mM GSH (red arrow). These results demonstrate that the PCDA-PEG-Biotin NPs are stable at blood concentration of GSH, but tend to lose their integrity at an intracellular concentration of GSH. This degradation behavior is thought to be attributed to the GSH concentration-dependent cleavage of the disulfide bond on the PCDA backbone of PCDA-PEG-Biotin molecule. Furthermore, the LC-HR-MS spectrum displayed two characteristic peaks of (M+ H)+ at about 369 and 370 m/z (Fig. 3E), revealing that the fragments degraded by 5 mM GSH were mainly CUR molecules (Mw = 368 Dalton). Therefore, it is expected that the PCDA-PEG-Biotin NPs will be stable for a considerably long period of time in the blood circulation, but will be readily degraded by high intracellular concentrations of GSH and then release the degraded fragment CUR once entering the targeted cancer cells 32,33 .
Meanwhile, the encapsulated DOX within the PCDA-PEG-Biotin NPs was also released responsively, as shown in Fig. 3F. In the absence of GSH, the DOX@PCDA-PEG-Biotin NPs almost did not release DOX. At a low concentration of GSH (10 μ M), a sustained DOX release behavior of the DOX@PCDA-PEG-Biotin NPs was observed, while at a relatively high concentration of GSH (5 mM), the release of DOX from the DOX@ PCDA-PEG-Biotin NPs became distinctly faster, owing to accelerated degradation and the disassembly of the hydrophobic PCDA zone as mentioned above.
Reversal Effects on MDR in MCF-7/ADR Cells. P-glycoprotein (P-gp), a product of the MDR1 gene, is a major adenosine triphosphate (ATP)-binding cassette (ABC) transporter which is able to function as an energy-dependent drug efflux pump and is linked to MDR in cancer cells 34 . ATP is a co-factor involved in drug efflux 35 . Herein, we investigated P-gp expression and ATP activity to evaluate reversal effects of the DOX@ PCDA-PEG-Biotin NPs on MDR in MCF-7/ADR human breast carcinoma cells. Compared with blank control, free DOX inhibited ATP activity of MCF-7/ADR cells intensively ( Fig. 4B) but also slightly increased P-gp overexpression (Fig. 4A), so that the intracellular accumulation of DOX was very limited (Fig. 4C,D), presumably due to a powerful MDR effect. Compared with free DOX, the combination of CUR with DOX not only inhibits ATP activity of MCF-7/ADR cells ( Fig. 4B) but also suppresses P-gp expression (Fig. 4A), which leads to slightly increased intracellular accumulation and retention of DOX (Fig. 4C,D), confirming the chemosensitization of CUR. In contrast, the DOX@PCDA-PEG NPs which can co-deliver CUR and DOX inhibited the overexpression of P-gp more noticeably. Therefore, the intracellular accumulation and retention amounts of DOX were enhanced greatly (Fig. 4C,D). The DOX@PCDA-PEG-Biotin NPs further suppressed P-gp expression and APT activity, and improved the intracellular DOX accumulation and retention, presumably owing to the active targeting effect of conjugated biotin. Figure 4E schematically illustrates the reversal effects of the DOX@PCDA-PEG-Biotin NPs on MDR in MCF-7/ADR cells, involving reduction of P-gp expression, inhibition of ATP activity and increase of intracellular drug accumulation.

Antitumor effects on MCF-7/ADR cells.
We have determined the IC50 values (drug concentrations which reduce cell viability by 50%) of DOX and its formulations against multidrug resistant MCF-7/ADR cells in order to assess their in vitro antitumor effects. As indicated in Table 1, the IC50 values of DOX for MCF-7 and MCF-7/ADR cells were measured to be 0.21 μ g/mL and 44.41 μ g/mL, respectively. The drug resistance index (DRI, the ratio of IC 50 values between MCF-7/ADR and MCF-7 cells) of MCF-7/ADR cells was calculated to be 211.48, indicating that cultured MCF-7/ADR cells were highly resistant to DOX. The cytotoxicities of CUR towards MCF-7 and MCF-7/ADR cells were relatively low as their IC 50 values were determined to be 19.52 μ g/mL and 51.76 μ g/mL, respectively ( Figure S6). However, DOX + CUR (0.8/1.0 (w/w)) had a much higher cytotoxicity against MCF-7/ADR cells (IC 50 = 14.29 μ g/mL) than DOX or CUR alone, owing to the chmosensitization of CUR as mentioned above. The reversal index (RI, the DRI of DOX /the DRI of DOX's formulation) of DOX + CUR was calculated to be 3.10 ( Table 1). The cytotoxicity of the PCDA-PEG-Biotin NPs against MCF-7 and MCF-7/ADR cells was dependent on their particle concentrations, but was still very limited at particle concentrations up to 160 μ g/mL ( Figure S8). However after combining/loading with DOX, the DOX@PCDA-PEG-Biotin NPs exhibited remarkably enhanced cytotoxicity against MCF-7/ADR cells as their IC 50 value was considerably lowered to 1.88 μ g/mL and their RI was as high as 23.62 (Table 1). In vivo tumor targeting and anti-MDR tumor efficacy. To evaluate the tumor targeting capability of the DOX@PCDA-PEG-Biotin NPs, the in vivo biodistribution of 1,1′ -dioctadecyl-3,3,3′ ,3′ -tetramethylindotricarbocyanine iodide (DiR)-labeled DOX@PCDA-PEG-Biotin NPs intravenously administered into the MCF-7/ADR tumor-bearing nude mice was investigated by a non-invasive near-infrared (NIR) optical imaging technique. The DiR-labeled DOX@PCDA-PEG-Biotin NPs presented a stronger fluorescence signal in the tumor region over a short time (6 h) compared with free DiR (Fig. 5A). This means that the amount of DiR-labeled DOX@PCDA-PEG-Biotin NPs in the tumor was higher than that of free DiR. Compared with the normal tissues, stronger fluorescence signals were found at the tumor site within 24 h post injection of the DiR-labeled DOX@PCDA-PEG-Biotin NPs, reflecting a notable tumor targeting effect of the DOX@ PCDA-PEG-Biotin NPs. At 24 h post injection, the mice were immediately euthanized, and several organs and tissues including tumor, heart, liver, spleen, lung and kidneys were harvested for ex vivo imaging. Compared with free DiR, the DiR-labeled DOX@PCDA-PEG-Biotin NPs had a much higher accumulation at the site of tumor site. These results clearly demonstrated an in vivo targeting ability of the DOX@PCDA-PEG-Biotin NPs.
Next, the antitumor effect of the DOX@PCDA-PEG-Biotin NPs was evaluated in MCF-7/ADR tumor xenograft model. As shown in Fig. 5, compared with free DOX, the DOX@PCDA-PEG-Biotin NPs presented a remarkably higher inhibition effect towards tumor growth, which is consistent with their in vitro antimultidrug resistant cancer effect. In addition, the body weights of the mice treated with DOX@PCDA-PEG-Biotin NPs, free DOX and saline had no significant change (Fig. 5C).
Histological examination by the hematoxylin and eosin (H&E) staining method showed that the treatment of the MCF-7/ADR tumor-bearing nude mice with the DOX@PCDA-PEG-Biotin NPs led to a great reduction in cancer cell density in the tumor tissue (Fig. 6). Moreover, the in situ Ki67 and TUNEL assays indicated that the treatment with DOX@PCDA-PEG-Biotin NPs significantly inhibited cancer cell proliferation and induced apoptosis in the tumor tissue (Fig. 6). Furthermore, the in situ P-gp assays showed that DOX@PCDA-PEG-Biotin NPs had greatest inhibition to P-gp expression in the tumor tissue as compared to other groups. Collectively, these results confirmed that the DOX@PCDA-PEG-Biotin NPs efficiently accumulated at the tumor site and thereby achieved an optimal antitumor efficacy in vivo.

Conclusion
In summary, we have developed a self-assembled polycurcumin nanoparticle, which can be used as both a pre-chemosensitizer and a nano-carrier of hydrophobic anticancer drug, by incorporating CUR into an intracellular cleavable polymer backbone. This nanoparticle shows GSH levels-dependent degradation and drug co-release properties, which are selectively degraded into CUR and simultaneously release the encapsulated DOX at intracellular GSH levels for enhanced biological specificity and therapeutic efficacy of multidrug resistant cancer. The hydrophilic outer shell of biotin-PEG offers effective protection against recognition and uptake by the reticuloendothelial system, allowing for stealth-shielding and long circulation, as well as biotin receptor-mediated tumor targeting. Furthermore, the polycurcumin nanoparticles could effectively reverse DOX resistance through suppressing P-gp expression and ATP activity, leading to enhanced cellular uptake of DOX and the reduced drug efflux in multidrug resistant MCF-7/ADR cells. Here DOX was incorporated into the hydrophobic polycurcumin inner core; we expected that other hydrophobic drugs could also be loaded into the polycurcumin domain through their hydrophobic interaction and be released by intracellular level of GSH triggered polycurcumin degradation.

Experimental Section
Materials. 3  In vitro GSH-triggered DOX release. The DOX release tests were carried out by a dialysis method. For example, DOX@PCDA-PEG-Biotin NPs (0.5 mL) were added into a dialysis bag with a molecular weight cut-off of 3500 Da) against 5 mL of the phosphate buffer solution (pH 7.4, 1% Tween-80) containing different concentrations of GSH, and gently shaken at 37 °C in a shaker at 120 rpm. At predetermined time intervals, the total buffer solution was withdrawn, followed by replacing with 5 ml of fresh buffer solution with the same GSH concentration. The fluorescence intensity of Dox was measured by a F-7000 fluorescence spectrophotometer.
After culture for 24 h, the cells were exposed to the DOX solution and the DOX-loaded NPs with different concentrations of DOX for 48 h, followed by adding 0.1 ml MTT (0.5 mg/ml). After 4 h of incubation, the culture medium was then removed and the cells were mixed with 0.1 ml of dimethyl sulphoxide. The absorbance was measured at a test wavelength of 570 nm by an ELISA plate reader (Varioskan Flash).
Flow cytometry. The fluorescence intensity of DOX in cells was measured by flow cytometry (FACS Calibur, BD, USA) and analyzed with CellQuest software through fluorescence channel2 (FL2).
P-gp and ATP assays. P-gp assay: MCF-7/ADR cells (2 × 105 cell/well) were seeded in 6-well plates. After culture for 24 h, the cells were exposed to the DOX solution (5 μ g/mL) and the DOX solution with 5 μ g/mL DOX and a CUR or its equivalent concentration of 6 μ g/mL for 48 h at 37 °C. Then the cells were trypsinized, collected, and resuspended in 0.1 mL of PBS (pH 7.4) for flow cytometry analysis. PE-conjugated mouse anti-human monoclonal antibody against P-gp was use to label cells according to the manufacture's instruction, and the nonspecific labeling was corrected by its isotype control. ATP assay: MCF-7/ADR cells (5 × 104 cells/well) were seeded in 24-well plates. After culturing for 24 h, the cells were exposed to the DOX solution (5 μ g/mL) and the DOX solution with 5 μ g/mL DOX and a CUR or its equivalent concentration of 6 μ g/mL for 48 h at 37 °C. Then the cells were washed with ice-cold PBS three times and lysed with ATP lysis buffer for ATP assay. The intracellular ATP levels were determined using a luciferin/luciferase assay according to the protocol of the ATP assay kit (Beyotime,China). In vivo imaging study. When the tumors reached to ~500 mm 3 , the mice were intravenously injected by DiR-labeled PCDA-PEG-Biotin NPs and free DiR (10 μ g), respectively. Images were taken on the IVIS-RLumina II imaging system (Caliper, USA, excitation: 748 nm, emmision:780 nm) at 2, 6 and 24 h post injection. After the 24 h scanning, the mice were euthanized. The tumors as well as major organs were harvested and subjected for ex vivo imaging.

Animals and tumor xenograft models.
In vivo antitumor efficacy. The tumor-bearing mice were weighed and randomly divided into different groups when the tumor volume reached to 100 mm 3 . From Day 0, the mice were intravenously injected with DOX solution (2 mg/kg), DOX@PCDA-PEG-Biotin NPs (2 mg/kg) and saline as a negative control once a week for 8 weeks, and meanwhile the body weight and the tumor size was measured. At Day 2 after the last drug administration, the mice were euthanized, and the tumor as well as the organs (heart, liver, spleen, lung, kidney) were collected and, weighed, washed with saline thrice and fixed in the 10% neutral-buffered formalin. For the hematoxylin and eosin staining, the formalin-fixed tumors were embedded in paraffin blocks and visualized by Olympus BX 51 microscope. For the TUNEL apoptosis staining, the fixed tumor sections were stained by the in Situ Cell Death Detection Kit (Roche Applied Science) according to the manufacturer protocol. The Ki-67 staining was conducted by using the labeled streptavidin-biotin method. The stained tumor slides were observed by Olympus BX 51 microscope. The inhibition rate of tumor growth (IRT) was calculated according to the following formula: IRT = 100% × (mean tumor weight of a control group − mean tumor weight of a treatment group)/mean tumor weight of a control group. The IRT values are listed in Table S3.